Source code for pysces.PyscesModel

"""
PySCeS - Python Simulator for Cellular Systems (http://pysces.sourceforge.net)

Copyright (C) 2004-2016 B.G. Olivier, J.M. Rohwer, J.-H.S Hofmeyr all rights reserved,

Brett G. Olivier (bgoli@users.sourceforge.net)
Triple-J Group for Molecular Cell Physiology
Stellenbosch University, South Africa.

Permission to use, modify, and distribute this software is given under the
terms of the PySceS (BSD style) license. See LICENSE.txt that came with
this distribution for specifics.

NO WARRANTY IS EXPRESSED OR IMPLIED.  USE AT YOUR OWN RISK.
Brett G. Olivier
"""

"""
FUTURE TODO:
Parameter elasticities wrt the compartments
"""

from pysces.version import __version__
__doc__ = '''
            PyscesModel
            -----------

            This module contains the core PySCeS classes which
            create the model and associated data objects

            '''
import os, copy, gc, time
import math, operator, re
import pprint, cPickle, cStringIO
import numpy
import scipy
import scipy.linalg
ScipyDerivative = None
HAVE_SCIPY_DERIV = False
try:
    ScipyDerivative = scipy.derivative
    HAVE_SCIPY_DERIV = True
except AttributeError:
    try:
        import scipy.misc
        ScipyDerivative = scipy.misc.derivative
        HAVE_SCIPY_DERIV = True
    except ImportError, AttributeError:
        pass
if not HAVE_SCIPY_DERIV:
    raise RuntimeError, '\nSciPy derivative function not available'

from getpass import getuser

from pysces import model_dir as MODEL_DIR
from pysces import output_dir as OUTPUT_DIR
from pysces import install_dir as INSTALL_DIR
from pysces import nleq2
from pysces import nleq2_switch
from pysces import pitcon
from pysces import plt, gplt
from pysces import PyscesStoich
from pysces import PyscesParse
from pysces import __SILENT_START__
from pysces.PyscesScan import Scanner
from pysces.core2.InfixParser import MyInfixParser
from pysces import SED
interface = None

# Scipy version check
if int(scipy.version.version.split('.')[1]) < 6:
    print '\nINFO: Your version of SciPy (' +  scipy.version.version + ') might be too old\n\tVersion 0.3.x or newer is strongly recommended\n'
else:
    if not __SILENT_START__:
        print 'You are using NumPy (%s) with SciPy (%s)' % (numpy.__version__, scipy.__version__)

_HAVE_PYSUNDIALS = False
_PYSUNDIALS_LOAD_ERROR = ''
try:
    from pysundials import cvode
    print 'PySundials available'
    _HAVE_PYSUNDIALS = True
except Exception, ex:
    _PYSUNDIALS_LOAD_ERROR = '%s' % ex
    _HAVE_PYSUNDIALS = False

# We now welcome a prototype StomPy interface to PySCeS which will provide expanding stochastic simulation support.


# for future fun
_HAVE_VPYTHON = False
_VPYTHON_LOAD_ERROR = ''
##  try:
    ##  import visual
    ##  _HAVE_VPYTHON = True
    ##  vpyscene = visual.display(x=150, y=50, title='PySCeS '+__version__+' - C Brett G. Olivier, Stellenbosch 2006', width=640, height=480,\
    ##  center=(0,0,0), background=(0,0,0))
    ##  vpyscene.select()
    ##  # vpyscene.autocenter = True
    ##  l1 = visual.label(pos=(0,0,0), text="Welcome to the PySCeS Visualisation Terminal", color=visual.color.red, box=0)
    ##  l2 = visual.label(pos=(0,0.5,0.5), text="It just works!", color=visual.color.green, box=0)
##  except Exception, ex:
    ##  _VPYTHON_LOAD_ERROR = '%s' % ex
    ##  _HAVE_VPYTHON = False

__psyco_active__ = 0
##  try:
    ##  import psyco
    ##  psyco.profile()
    ##  __psyco_active__ = 1
    ##  print 'PySCeS Model module is now PsycoActive!'
##  except:
    ##  __psyco_active__ = 0

# machine specs
mach_spec = numpy.MachAr()
# grab a parser
pscParser = PyscesParse.PySCeSParser(debug=0)

[docs]def chkpsc(File): """ chkpsc(File) Chekc whether the filename "File" has a '.psc' extension and adds one if not. Arguments: File: filename string """ try: if File[-4:] == '.psc': pass else: print 'Assuming extension is .psc' File += '.psc' except: print 'Chkpsc error' return File
[docs]def chkmdir(): """ chkmdir() Import and grab pysces.model_dir Arguments: None """ ## from pysces import model_dir as MODEL_DIR pass
[docs]class BagOfStuff(object): """ A collection of attributes defined by row and column lists used by Response coefficients etc matrix is an array of values while row/col are lists of row colummn name strings """ matrix = None row = None col = None def __init__(self, matrix, row, col): "Load attributes from arrays" assert len(row) == matrix.shape[0], "\nRow dimension mismatch" assert len(col) == matrix.shape[1], "\nCol dimension mismatch" self.matrix = matrix self.row = list(row) self.col = list(col)
[docs] def load(self): for r in range(self.matrix.shape[0]): for c in range(self.matrix.shape[1]): setattr(self, str(self.row[r]) + '_' + str(self.col[c]), self.matrix[r,c])
[docs] def get(self, attr1, attr2): "Returns a single attribute \"attr1_attr2\" or None" try: return getattr(self, attr1 + '_' + attr2) except Exception, ex: print ex return None
[docs] def select(self, attr, search='a'): """Return a dictionary of <attr>_<name>, <name>_<attr> : val or {} if none If attr exists as an index for both left and right attr then: search='a' : both left and right attributes (default) search='l' : left attributes only search='r' : right attributes""" output = {} if attr in self.row and attr in self.col: if search in ['a','l']: row = self.row.index(attr) for col in range(self.matrix.shape[1]): output.setdefault(attr + '_' + self.col[col], self.matrix[row,col]) if search in ['a','r']: col = self.col.index(attr) for row in range(self.matrix.shape[0]): output.setdefault(self.row[row] + '_' + attr, self.matrix[row,col]) elif attr in self.row: row = self.row.index(attr) for col in range(self.matrix.shape[1]): output.setdefault(attr + '_' + self.col[col], self.matrix[row,col]) elif attr in self.col: col = self.col.index(attr) for row in range(self.matrix.shape[0]): output.setdefault(self.row[row] + '_' + attr, self.matrix[row,col]) return output
[docs] def list(self): """ Return all attributes as a attr:val dictionary """ output = {} for row in range(self.matrix.shape[0]): for col in range(self.matrix.shape[1]): output.setdefault(self.row[row] + '_' + self.col[col], self.matrix[row,col]) return output
[docs] def listAllOrdered(self, order='descending', absolute=True): """ Return an ordered list of (attr, value) tuples - *order* [default='descending'] the order to return as: 'descending' or 'ascending' - *absolute* [default=True] use the absolute value """ if order != 'ascending': order = True else: order = False output_v = [] output_n = [] for row in range(self.matrix.shape[0]): for col in range(self.matrix.shape[1]): output_v.append(self.matrix[row,col]) output_n.append(self.row[row] + '_' + self.col[col]) if absolute: new_idx = numpy.argsort([abs(v) for v in output_v]) else: new_idx = numpy.argsort(output_v) output = [] if order: new_idx = new_idx.tolist() new_idx.reverse() for i_ in new_idx: output.append((output_n[i_], output_v[i_])) return output
[docs]class ScanDataObj(object): """ New class used to store parameter scan data (uses StateDataObj) """ species = None fluxes = None rules = None xdata = None mod_data = None flux_labels = None species_labels = None rules_labels = None xdata_labels = None mod_data_labels = None invalid_states = None parameters = None parameter_labels = None HAS_FLUXES = False HAS_SPECIES = False HAS_RULES = False HAS_XDATA = False HAS_MOD_DATA = False HAS_SET_LABELS = False ALL_VALID = True OPEN = True def __init__(self, par_label): self.species = [] self.fluxes = [] self.rules = [] self.xdata = [] self.mod_data = [] self.invalid_states = [] self.parameters = [] if isinstance(par_label, list): self.parameter_labels = par_label else: self.parameter_labels = [par_label]
[docs] def setLabels(self, ssdata): if ssdata.HAS_SPECIES: self.species_labels = ssdata.species_labels self.HAS_SPECIES = True if ssdata.HAS_FLUXES: self.flux_labels = ssdata.flux_labels self.HAS_FLUXES = True if ssdata.HAS_RULES: self.rules_labels = ssdata.rules_labels self.HAS_RULES = True if ssdata.HAS_XDATA: self.xdata_labels = ssdata.xdata_labels self.HAS_XDATA = True self.HAS_SET_LABELS = True
[docs] def addPoint(self, ipar, ssdata): """takes a list/array of input parameter values and the associated ssdata object""" assert self.OPEN, '\nScan has been finalised no new data may be added' if not self.HAS_SET_LABELS: self.setLabels(ssdata) if hasattr(ipar, '__iter__'): self.parameters.append(ipar) else: self.parameters.append([ipar]) if self.HAS_SPECIES: self.species.append(ssdata.getSpecies()) if self.HAS_FLUXES: self.fluxes.append(ssdata.getFluxes()) if self.HAS_RULES: self.rules.append(ssdata.getRules()) if self.HAS_XDATA: self.xdata.append(ssdata.getXData()) if not ssdata.IS_VALID: self.ALL_VALID = False self.invalid_states.append(ipar)
[docs] def addModData(self, mod, *args): if not self.HAS_MOD_DATA: self.HAS_MOD_DATA = True self.mod_data_labels = [attr for attr in args if hasattr(mod, attr)] self.mod_data.append(tuple([getattr(mod, attr) for attr in self.mod_data_labels]))
[docs] def closeScan(self): print '\nINFO: closing scan no new data may be added.' if self.HAS_SPECIES: self.species = numpy.array(self.species) if self.HAS_FLUXES: self.fluxes = numpy.array(self.fluxes) if self.HAS_RULES: self.rules = numpy.array(self.rules) if self.HAS_XDATA: self.xdata = numpy.array(self.xdata) if self.HAS_MOD_DATA: self.mod_data = numpy.array(self.mod_data) self.parameters = numpy.array(self.parameters) self.OPEN = False
[docs] def getSpecies(self, lbls=False): if self.OPEN: self.closeScan() output = None labels = None if self.HAS_SPECIES: output = numpy.hstack((self.parameters, self.species)) labels = self.parameter_labels+self.species_labels if lbls: return output, labels else: return output
[docs] def getFluxes(self, lbls=False): if self.OPEN: self.closeScan() output = None labels = None if self.HAS_FLUXES: output = numpy.hstack((self.parameters, self.fluxes)) labels = self.parameter_labels+self.flux_labels if lbls: return output, labels else: return output
[docs] def getRules(self, lbls=False): if self.OPEN: self.closeScan() output = None labels = None if self.HAS_RULES: output = numpy.hstack((self.parameters, self.rules)) labels = self.parameter_labels+self.rules_labels if lbls: return output, labels else: return output
[docs] def getXData(self, lbls=False): if self.OPEN: self.closeScan() output = None labels = None if self.HAS_XDATA: output = numpy.hstack((self.parameters, self.xdata)) labels = self.parameter_labels+self.xdata_labels if lbls: return output, labels else: return output
[docs] def getModData(self, lbls=False): if self.OPEN: self.closeScan() output = None labels = None if self.HAS_MOD_DATA: output = numpy.hstack((self.parameters, self.mod_data)) labels = self.parameter_labels+self.mod_data_labels if lbls: return output, labels else: return output
[docs] def getAllScanData(self, lbls=False): if self.OPEN: self.closeScan() output = self.parameters labels = self.parameter_labels if self.HAS_SPECIES: output = numpy.hstack((output, self.species)) labels = labels+self.species_labels if self.HAS_FLUXES: output = numpy.hstack((output, self.fluxes)) labels = labels+self.flux_labels if self.HAS_RULES: output = numpy.hstack((output, self.rules)) labels = labels+self.rules_labels if self.HAS_XDATA: output = numpy.hstack((output, self.xdata)) labels = labels+self.xdata_labels if self.HAS_MOD_DATA: output = numpy.hstack((output, self.mod_data)) labels = labels+self.mod_data_labels if lbls: return output, labels else: return output
[docs] def getScanData(self, *args, **kwargs): """getScanData(\*args) feed this method species/flux/rule/mod labels and it will return an array of [parameter(s), sp1, f1, ....] """ if kwargs.has_key('lbls'): lbls = kwargs['lbls'] else: lbls = False output = self.parameters lout = self.parameter_labels for roc in args: if self.HAS_SPECIES and roc in self.species_labels: lout.append(roc) output = numpy.hstack((output, self.species.take([self.species_labels.index(roc)], axis=-1))) if self.HAS_FLUXES and roc in self.flux_labels: lout.append(roc) output = numpy.hstack((output, self.fluxes.take([self.flux_labels.index(roc)], axis=-1))) if self.HAS_RULES and roc in self.rules_labels: lout.append(roc) output = numpy.hstack((output, self.rules.take([self.rules_labels.index(roc)], axis=-1))) if self.HAS_XDATA and roc in self.xdata_labels: lout.append(roc) output = numpy.hstack((output, self.xdata.take([self.xdata_labels.index(roc)], axis=-1))) if self.HAS_MOD_DATA and roc in self.mod_data_labels: lout.append(roc) output = numpy.hstack((output, self.mod_data.take([self.mod_data_labels.index(roc)], axis=-1))) if not lbls: return output else: return output, lout
[docs]class StateDataObj(object): """ New class used to store steady-state data. """ fluxes = None species = None rules = None xdata = None flux_labels = None species_labels = None rules_labels = None xdata_labels = None HAS_FLUXES = False HAS_SPECIES = False HAS_RULES = False HAS_XDATA = False IS_VALID = True ## def setLabels(self, species=None, fluxes=None, rules=None): ## """set the species, rate and rule label lists""" ## if species != None: ## self.species_labels = species ## if fluxes != None: ## self.flux_labels = fluxes ## if rules != None: ## self.rules_labels = rules
[docs] def setSpecies(self, species, lbls=None): """Set the species array""" self.species = species self.HAS_SPECIES = True if lbls != None: self.species_labels = lbls for s in range(len(self.species_labels)): setattr(self, self.species_labels[s], self.species[s])
[docs] def setFluxes(self, fluxes, lbls=None): """set the flux array""" self.fluxes = fluxes self.HAS_FLUXES = True if lbls != None: self.flux_labels = lbls for f in range(len(self.flux_labels)): setattr(self, self.flux_labels[f], self.fluxes[f])
[docs] def setRules(self, rules, lbls=None): """Set the results of rate rules""" self.rules = rules self.HAS_RULES = True if lbls != None: self.rules_labels = lbls for r in range(len(self.rules_labels)): setattr(self, self.rules_labels[r], self.rules[r])
[docs] def setXData(self, xdata, lbls=None): """Sets extra simulation data""" self.xdata = xdata self.HAS_XDATA = True if lbls != None: self.xdata_labels = lbls for x in range(len(self.xdata_labels)): setattr(self, self.xdata_labels[x], self.xdata[x])
[docs] def getSpecies(self, lbls=False): """return species array""" output = None if self.HAS_SPECIES: output = self.species if not lbls: return output else: return output, self.species_labels
[docs] def getFluxes(self, lbls=False): """return flux array""" output = None if self.HAS_FLUXES: output = self.fluxes if not lbls: return output else: return output, self.flux_labels
[docs] def getRules(self, lbls=False): """Return rule array""" output = None if self.HAS_RULES: output = self.rules if not lbls: return output else: return output, self.rules_labels
[docs] def getXData(self, lbls=False): """Return xdata array""" output = None if self.HAS_XDATA: output = self.xdata if not lbls: return output else: return output, self.xdata_labels
[docs] def getAllStateData(self, lbls=False): """ Return all available data as species+fluxes+rules if lbls=True returns (array,labels) else just array """ labels = [] output = None if self.HAS_SPECIES: output = self.species labels += self.species_labels if self.HAS_FLUXES: if output == None: output = self.fluxes else: output = numpy.hstack((output, self.fluxes)) labels += self.flux_labels if self.HAS_RULES: if output == None: output = self.rules else: output = numpy.hstack((output, self.rules)) labels += self.rules_labels if self.HAS_XDATA: if output == None: output = self.xdata else: output = numpy.hstack((output, self.xdata)) labels += self.xdata_labels if not lbls: return output else: return output, labels
[docs] def getStateData(self, *args, **kwargs): """getSimData(\*args) feed this method species/rate labels and it will return an array of [time, sp1, r1, ....] """ if kwargs.has_key('lbls'): lbls = kwargs['lbls'] else: lbls = False lout = [] output = [] for roc in args: if self.HAS_SPECIES and roc in self.species_labels: lout.append(roc) output.append(self.species[self.species_labels.index(roc)]) elif self.HAS_FLUXES and roc in self.flux_labels: lout.append(roc) output.append(self.fluxes[self.flux_labels.index(roc)]) elif self.HAS_RULES and roc in self.rules_labels: lout.append(roc) output.append(self.rules[self.rules_labels.index(roc)]) elif self.HAS_XDATA and roc in self.xdata_labels: lout.append(roc) output.append(self.xdata[self.xdata_labels.index(roc)]) else: print 'I don\'t have an attribute %s ... ignoring.' % roc if not lbls: return output else: return numpy.array(output), lout
[docs]class IntegrationDataObj(object): """ This class is specifically designed to store the results of a time simulation It has methods for setting the Time, Labels, Species and Rate data and getting Time, Species and Rate (including time) arrays. However, of more use: - getOutput(\*args) feed this method species/rate labels and it will return an array of [time, sp1, r1, ....] - getDataAtTime(time) the data generated at time point "time". - getDataInTimeInterval(time, bounds=None) more intelligent version of the above returns an array of all data points where: time-bounds <= time <= time+bounds """ time = None rates = None species = None rules = None xdata = None time_label = 'Time' rate_labels = None species_labels = None rules_labels = None xdata_labels = None HAS_SPECIES = False HAS_RATES = False HAS_RULES = False HAS_TIME = False HAS_XDATA = False IS_VALID = True TYPE_INFO = 'Deterministic'
[docs] def setLabels(self, species=None, rates=None, rules=None): """set the species, rate and rule label lists""" if species != None: self.species_labels = species if rates != None: self.rate_labels = rates if rules != None: self.rules_labels = rules
[docs] def setTime(self, time, lbl=None): """Set the time vector""" self.time = time.reshape(len(time), 1) self.HAS_TIME = True if lbl != None: self.time_label = lbl
[docs] def setSpecies(self, species, lbls=None): """Set the species array""" self.species = species self.HAS_SPECIES = True if lbls != None: self.species_labels = lbls
[docs] def setRates(self, rates, lbls=None): """set the rate array""" self.rates = rates self.HAS_RATES = True if lbls != None: self.rate_labels = lbls
[docs] def setRules(self, rules, lbls=None): """Set the results of rate rules""" self.rules = rules self.HAS_RULES = True if lbls != None: self.rules_labels = lbls
[docs] def setXData(self, xdata, lbls=None): """Sets extra simulation data""" self.xdata = xdata self.HAS_XDATA = True if lbls != None: self.xdata_labels = lbls
[docs] def getTime(self, lbls=False): """return the time vector""" output = None if self.HAS_TIME: output = self.time.reshape(len(self.time),) if not lbls: return output else: return output, [self.time_label]
[docs] def getSpecies(self, lbls=False): """return time+species array""" output = None if self.HAS_SPECIES: output = numpy.hstack((self.time, self.species)) labels = [self.time_label]+self.species_labels else: output = self.time labels = [self.time_label] if not lbls: return output else: return output, labels
[docs] def getRates(self, lbls=False): """return time+rate array""" output = None if self.HAS_RATES: output = numpy.hstack((self.time, self.rates)) labels = [self.time_label]+self.rate_labels else: output = self.time labels = [self.time_label] if not lbls: return output else: return output, labels
[docs] def getRules(self, lbls=False): """Return time+rule array""" ## assert self.rules != None, "\nNo rules" output = None if self.HAS_RULES: output = numpy.hstack((self.time, self.rules)) labels = [self.time_label]+self.rules_labels else: output = self.time labels = [self.time_label] if not lbls: return output else: return output, labels
[docs] def getXData(self, lbls=False): """Return time+xdata array""" ## assert self.rules != None, "\nNo rules" output = None if self.HAS_XDATA: output = numpy.hstack((self.time, self.xdata)) labels = [self.time_label]+self.xdata_labels else: output = self.time labels = [self.time_label] if not lbls: return output else: return output, labels
[docs] def getDataAtTime(self, time): """Return all data generated at "time" """ #TODO add rate rule data t = None sp = None ra = None ru = None xd = None temp_t = self.time.reshape(len(self.time),) for tt in range(len(temp_t)): if temp_t[tt] == time: t = tt if self.HAS_SPECIES: sp = self.species.take([tt], axis=0) if self.HAS_RATES: ra = self.rates.take([tt], axis=0) if self.HAS_RULES: ru = self.rules.take([tt], axis=0) if self.HAS_XDATA: xd = self.xdata.take([tt], axis=0) break output = None if t is not None: output = numpy.array([[temp_t[t]]]) if sp is not None: output = numpy.hstack((output,sp)) if ra is not None: output = numpy.hstack((output,ra)) if ru is not None: output = numpy.hstack((output,ru)) if xd is not None: output = numpy.hstack((output,xd)) return output
[docs] def getDataInTimeInterval(self, time, bounds=None): """ getDataInTimeInterval(time, bounds=None) returns an array of all data points where: time-bounds <= time <= time+bounds where bound defaults to stepsize """ #TODO add rate rule data temp_t = self.time.reshape(len(self.time),) if bounds == None: bounds = temp_t[1] - temp_t[0] c1 = (temp_t >= time-bounds) c2 = (temp_t <= time+bounds) print 'Searching (%s:%s:%s)' % (time-bounds, time, time+bounds) t = [] sp = None ra = None for tt in range(len(c1)): if c1[tt] and c2[tt]: t.append(tt) output = None if len(t) > 0: output = self.time.take(t) output = output.reshape(len(output),1) if self.HAS_SPECIES and self.HAS_TIME: output = numpy.hstack((output, self.species.take(t, axis=0))) if self.HAS_RATES: output = numpy.hstack((output, self.rates.take(t, axis=0))) if self.HAS_RULES: output = numpy.hstack((output, self.rules.take(t, axis=0))) if self.HAS_XDATA: output = numpy.hstack((output, self.xdata.take(t, axis=0))) return output
[docs] def getOutput(self, *args, **kwargs): """ Old alias for getSimData() getOutput(\*args) feed this method species/rate labels and it will return an array of [time, sp1, r1, ....] """ return self.getSimData(*args, **kwargs)
[docs] def getAllSimData(self,lbls=False): """ Return all available data as time+species+rates+rules if lbls=True returns (array,lables) else just array """ labels = [self.time_label] if self.HAS_SPECIES and self.HAS_TIME: output = numpy.hstack((self.time, self.species)) labels += self.species_labels if self.HAS_RATES: output = numpy.hstack((output, self.rates)) labels +=self.rate_labels if self.HAS_RULES: output = numpy.hstack((output, self.rules)) labels += self.rules_labels if self.HAS_XDATA: output = numpy.hstack((output, self.xdata)) labels += self.xdata_labels if not lbls: return output else: return output, labels
[docs] def getSimData(self, *args, **kwargs): """getSimData(\*args) feed this method species/rate labels and it will return an array of [time, sp1, r1, ....] """ output = self.time ## print argimrgs if kwargs.has_key('lbls'): lbls = kwargs['lbls'] else: lbls = False lout = [self.time_label] for roc in args: if self.HAS_SPECIES and roc in self.species_labels: lout.append(roc) output = numpy.hstack((output, self.species.take([self.species_labels.index(roc)], axis=-1))) if self.HAS_RATES and roc in self.rate_labels: lout.append(roc) output = numpy.hstack((output, self.rates.take([self.rate_labels.index(roc)], axis=-1))) if self.HAS_RULES and roc in self.rules_labels: lout.append(roc) output = numpy.hstack((output, self.rules.take([self.rules_labels.index(roc)], axis=-1))) if self.HAS_XDATA and roc in self.xdata_labels: lout.append(roc) output = numpy.hstack((output, self.xdata.take([self.xdata_labels.index(roc)], axis=-1))) if not lbls: return output else: return output, lout # this must stay in sync with core2
[docs]class NewCoreBase(object): """ Core2 base class, needed here as we use Core2 derived classes in PySCes """ name = None __DEBUG__ = False
[docs] def getName(self): return self.name
[docs] def setName(self,name): self.name = name
[docs] def get(self, attr): """Return an attribute whose name is str(attr)""" return self.__getattribute__(attr) # this must stay in sync with core2
[docs]class NumberBase(NewCoreBase): """ Derived Core2 number class. """ value = None value_initial = None def __call__(self): return self.value
[docs] def getValue(self): return self.value
[docs] def setValue(self, v): self.value = v # Finally killed my lambda functions - brett07
[docs]class ReactionObj(NewCoreBase): """ Defines a reaction with a KineticLaw *kl8, *formula* and *name* bound to a model instance, *mod*. """ formula = None mod = None rate = None xcode = None code_string = None symbols = None compartment = None piecewises = None def __init__(self, mod, name, kl, klrepl='self.'): """ mod : model instance name = reaction name kl = kinetic law klrepl = replacement string for KineticLaw """ self.mod = mod self.name = name self.setKineticLaw(kl, klrepl='self.') def __call__(self, *args): exec(self.xcode) return self.rate
[docs] def setKineticLaw(self, kl, klrepl='self.'): InfixParser.setNameStr('self.mod.', '') InfixParser.parse(kl.replace(klrepl, '')) self.symbols = InfixParser.names self.piecewises = InfixParser.piecewises formula = InfixParser.output # this code has to stay together # for pw in InfixParser.piecewises: formula = formula.replace('self.mod.%s' % pw, 'self.mod.%s()' % pw) # this code has to stay together # self.code_string = 'self.rate=%s' % formula self.xcode = compile(self.code_string, 'Req: %s' % self.name, 'exec') self.formula = kl.replace(klrepl, '').replace('numpy.', '').replace('math.', '').replace('operator.', '')
InfixParser = MyInfixParser() InfixParser.buildlexer() InfixParser.buildparser(debug=0, debugfile=os.path.join(OUTPUT_DIR, 'infix.dbg'),\ tabmodule=os.path.join(OUTPUT_DIR, 'infix_tabmodule')) InfixParser.setNameStr('self.', '') os.chdir(OUTPUT_DIR) # adapted from core2
[docs]class Function(NewCoreBase): """ Function class ported from Core2 to enable the use of functions in PySCeS. """ formula = None code_string = None xcode = None value = None symbols = None argsl = None mod = None piecewises = None functions = None def __init__(self, name, mod): self.name = name self.argsl = [] self.functions = [] self.mod = mod def __call__(self, *args): for ar in range(len(args)): self.__setattr__(self.argsl[ar], args[ar]) exec(self.xcode) return self.value
[docs] def setArg(self, var, value=None): self.__setattr__(var, value) self.argsl.append(var)
[docs] def addFormula(self, formula): self.formula = formula InfixParser.setNameStr('self.', '') InfixParser.SymbolReplacements = {'_TIME_':'mod._TIME_'} InfixParser.parse(formula) self.piecewises = InfixParser.piecewises self.symbols = InfixParser.names self.functions = InfixParser.functions self.code_string = 'self.value=%s' % InfixParser.output self.xcode = compile(self.code_string, 'Func: %s' % self.name, 'exec') # adapted from core2
[docs]class EventAssignment(NumberBase): """ Event assignments are actions that are triggered by an event. Ported from Core2 to build an event handling framework fro PySCeS """ variable = None symbols = None formula = None code_string = None xcode = None mod = None piecewises = None __DEBUG__ = False def __call__(self): setattr(self.mod, self.variable, self.value) if self.__DEBUG__: print '\tAssigning %s = %s' % (self.variable, self.value) return True def __init__(self, name, mod): self.setName(name) self.mod = mod
[docs] def setVariable(self, var): self.variable = var
[docs] def setFormula(self, formula): self.formula = formula InfixParser.setNameStr('self.mod.', '') InfixParser.SymbolReplacements = {'_TIME_':'self.mod._TIME_'} InfixParser.parse(formula) self.piecewises = InfixParser.piecewises self.symbols = InfixParser.names self.code_string = 'self.value=%s' % InfixParser.output self.xcode = compile(self.code_string, 'EvAs: %s' % self.name, 'exec') ## print '\t', self.name, self.code_string
[docs] def evaluateAssignment(self): exec(self.xcode) # adapted from core2
[docs]class Event(NewCoreBase): """ Event's have triggers and fire EventAssignments when required. Ported from Core2. """ trigger = None delay = 0.0 formula = None code_string = None xcode = None state0 = False state = False assignments = None _TIME_ = 0.0 _ASS_TIME_ = 0.0 _need_action = False symbols = None _time_symbol = None piecewises = None mod = None __DEBUG__ = True def __init__(self, name, mod): self.setName(name) self.assignments = [] self.mod = mod def __call__(self, time): self._TIME_ = time exec(self.xcode) ret = False if self.state0 and not self.state: self.state0 = self.state if not self.state0 and self.state: for ass in self.assignments: ass.evaluateAssignment() self.state0 = self.state self._need_action = True self._ASS_TIME_ = time + self.delay if self.__DEBUG__: print '\nevent %s is evaluating at %s' % (self.name, time) ret = True if self._need_action and self._TIME_ >= self._ASS_TIME_: for ass in self.assignments: ass() if self.__DEBUG__: print 'event %s is assigning at %s (delay=%s)' % (self.name, time, self.delay) self._need_action = False ret = True return ret
[docs] def setTrigger(self, formula, delay=0.0): self.formula = formula self.delay = delay InfixParser.setNameStr('self.mod.', '') ## print formula, delay if self._time_symbol != None: InfixParser.SymbolReplacements = {self._time_symbol : '_TIME_'} else: InfixParser.SymbolReplacements = {'_TIME_' : '_TIME_'} InfixParser.parse(formula) self.piecewises = InfixParser.piecewises self.symbols = InfixParser.names self.code_string = 'self.state=%s' % InfixParser.output self.xcode = compile(self.code_string, 'Ev: %s' % self.name, 'exec') ## print self.name, self.code_string
[docs] def setAssignment(self, var, formula): ass = EventAssignment(var, mod=self.mod) ass.setVariable(var) ass.setFormula(formula) self.assignments.append(ass) self.__setattr__('_'+var, ass)
[docs] def reset(self): self.state0 = False self.state = False self._TIME_ = 0.0 self._ASS_TIME_ = 0.0
[docs]class PieceWise(NewCoreBase): """ Generic piecewise class adapted from Core2 that generates a compiled Python code block that allows evaluation of arbitrary length piecewise functions. Piecewise statements should be defined in assignment rules as `piecewise(<Piece>, <Conditional>, <OtherValue>)` where there can be an arbitrary number of `<Piece>, <Conditional>` pairs. - *args* a dictionary of piecewise information generated by the InfixParser as InfixParser.piecewises """ name = None value = None formula = None code_string = None xcode = None _names = None _TIME_ = None def __init__(self, pwd, mod): pwd = pwd.copy() self.mod = mod if pwd['other'] != None: other = 'self.value = %s' % pwd.pop('other').replace('self.','self.mod.') else: other = 'pass' pwd.pop('other') InfixParser.setNameStr('self.mod.', '') self._names = [] if len(pwd.keys()) == 1: formula = pwd[0][0] InfixParser.parse(formula) for n in InfixParser.names: if n not in self._names and n != '_TIME_': self._names.append(n) formula = InfixParser.output thenStat = pwd[0][1].replace('self.','self.mod.') ## thenStat = pwd[0][1] ## InfixParser.setNameStr('self.mod.', '') ## InfixParser.parse(thenStat) ## thenStat = InfixParser.output self.code_string = 'if %s:\n self.value = %s\nelse:\n %s' %\ (formula, thenStat, other) self.formula = self.code_string.replace('self.','') else: formula = pwd[0][0] InfixParser.parse(formula) for n in InfixParser.names: if n not in self._names and n != '_TIME_': self._names.append(n) formula = InfixParser.output thenStat = pwd[0][1].replace('self.','self.mod.') ## thenStat = pwd[0][1] ## InfixParser.setNameStr('self.mod.', '') ## InfixParser.parse(thenStat) ## thenStat = InfixParser.output self.code_string = 'if %s:\n self.value = %s\n' % (formula, thenStat) pwd.pop(0) for p in pwd: formula = pwd[p][0] InfixParser.parse(formula) for n in InfixParser.names: if n not in self._names and n != '_TIME_': self._names.append(n) formula = InfixParser.output thenStat = pwd[p][1].replace('self.','self.mod.') ## thenStat = pwd[p][1] ## InfixParser.setNameStr('self.mod.', '') ## InfixParser.parse(thenStat) ## thenStat = InfixParser.output self.code_string += 'elif %s:\n self.value = %s\n' % (formula, thenStat) self.code_string += 'else:\n %s' % other self.formula = self.code_string.replace('self.','') self.xcode = compile(self.code_string, 'PieceWise','exec') def __call__(self): exec(self.xcode) return self.value
[docs]class PysMod(object): """ Create a model object and instantiate a PySCeS model so that it can be used for further analyses. PySCeS model descriptions can be loaded from files or strings (see the *loader* argument for details). - *File* the name of the PySCeS input file if not explicit a \*.psc extension is assumed. - *dir* if specified, the path to the input file otherwise the default PyscesModel directory (defined in the pys_config.ini file) is assumed. - *autoload* autoload the model, pre 0.7.1 call mod.doLoad(). (default=True) **new** - *loader* the default behaviour is to load PSC file, however, if this argument is set to 'string' an input file can be supplied as the *fString* argument (default='file') - *fString* a string containing a PySCeS model file (use with *loader='string'*) the *File* argument now sepcifies the new input file name. """ __version__ = __version__ __pysces_directory__ = INSTALL_DIR __settings__ = None #__STOMPY__ = None def __init__(self,File=None,dir=None,loader='file',fString=None,autoload=True): """ Create a model object and instantiate a PySCeS model so that it can be used for further analyses. PySCeS model descriptions can be loaded from files or strings (see the *loader* argument for details). - *File* the name of the PySCeS input file if not explicit a \*.psc extension is assumed. - *dir* if specified, the path to the input file otherwise the default PyscesModel directory (defined in the pys_config.ini file) is assumed. - *autoload* autoload the model, pre 0.7.1 call mod.doLoad(). (default=True) **new** - *loader* the default behaviour is to load PSC file, however, if this argument is set to 'string' an input file can be supplied as the *fString* argument (default='file') - *fString* a string containing a PySCeS model file (use with *loader='string'*) the *File* argument now sepcifies the new input file name. """ #if _HAVE_STOMPY: #self.__STOMPY__ = StomPyInterface(MODEL_DIR, OUTPUT_DIR) #print 'PySCeS/StochPy interface active' #else: #self.__STOMPY__ = None self.__settings__ = {} self.__settings__.update({'enable_deprecated_attr' : True}) if loader == 'file': self.LoadFromFile(File,dir) elif loader == 'string': self.LoadFromString(File,fString) else: self.LoadFromFile(File,dir) # stuff that needs to be done before initmodel self.__settings__['mode_substitute_assignment_rules'] = False self.__settings__['display_compartment_warnings'] = False self._TIME_ = 0.0 # this will be the built-in time self.piecewise_functions = [] self.__piecewises__ = {} self.__HAS_PIECEWISE__ = False if autoload: self.ModelLoad() self.__PSC_auto_load = True else: self.__PSC_auto_load = False
[docs] def ModelLoad(self,stoich_load=0): """ Load and instantiate a PySCeS model so that it can be used for further analyses. This function replaces the pre-0.7.1 doLoad() method. - *stoich_load* try to load a structural analysis saved with Stoichiometry_Save_Serial() (default=0) """ self.InitialiseInputFile() assert self.__parseOK, '\nError in input file, parsing could not complete' self.Stoichiometry_Analyse(override=0,load=stoich_load) # add new Style functions to model self.InitialiseFunctions() self.InitialiseCompartments() self.InitialiseRules() self.InitialiseEvents() self.InitialiseModel() self.InitialiseRuleChecks() self.InitialiseOldFunctions() # TODO replace this with initialisation functions
[docs] def doLoad(self,stoich_load=0): """ Load and instantiate a PySCeS model so that it can be used for further analyses. This function is being replaced by the ModelLoad() method. - *stoich_load* try to load a structural analysis saved with Stoichiometry_Save_Serial() (default=0) """ if not self.__PSC_auto_load: self.ModelLoad(stoich_load=stoich_load) else: print 'PySCeS now automatically loads the model on model object instantiation. If you do not want this behaviour pass the autoload=False argument to the constructor, if you really want to reload the model, run doLoad() again.\n' print 'Further calls to doLoad() will work as normal.' self.__PSC_auto_load = False
[docs] def LoadFromString(self,File=None,fString=None): """ Docstring required """ #grab model directory chkmdir() dir = MODEL_DIR # check for .psc extension File = chkpsc(File) print 'Using model directory: ' + dir if not os.path.isdir(os.path.join(MODEL_DIR,"orca")): os.mkdir(os.path.join(MODEL_DIR,"orca")) dir = os.path.join(MODEL_DIR,"orca") # write string to file try: outFile = file(os.path.join(dir,File),'w') outFile.write(fString) outFile.close() print 'Using file: ' + File if os.path.exists(os.path.join(dir,File)): print os.path.join(dir,File) + ' loading .....', self.ModelDir = dir self.ModelFile = File else: print os.path.join(dir,File) + ' does not exist' print 'Please set with ModelDir and ModelFile .....', self.ModelFile = 'None' self.ModelDir = dir except Exception, e: print e print os.path.join(dir,File) + ' does not exist please re-instantiate model .....', self.__settings__['display_debug'] = 0 self.ModelOutput = OUTPUT_DIR ## # Initialise elementary modes ## if os.sys.platform == 'win32': ## self.eModeExe_int = os.path.join(self.__metatool,'meta43_int.exe') ## self.eModeExe_dbl = os.path.join(self.__metatool,'meta43_double.exe') ## else: ## self.eModeExe_int = os.path.join(self.__metatool,'meta43_int') ## self.eModeExe_dbl = os.path.join(self.__metatool,'meta43_double') ## print 'Done.' # Initialize serial directory self.__settings__['serial_dir'] = os.path.join(self.ModelOutput,'pscdat') # Initialize stoichiometric precision self.__settings__['stoichiometric_analysis_fp_zero'] = mach_spec.eps*2.0e4 self.__settings__['stoichiometric_analysis_lu_precision'] = self.__settings__['stoichiometric_analysis_fp_zero'] self.__settings__['stoichiometric_analysis_gj_precision'] = self.__settings__['stoichiometric_analysis_lu_precision']*10.0
[docs] def LoadFromFile(self,File=None,dir=None): """ __init__(File=None,dir=None) Initialise a PySCeS model object with PSC file that can be found in optional directory. If a a filename is not supplied the pysces.model_dir directory contents is displayed and the model name can be entered at the promp (<ctrl>+C exits the loading process). Arguments: File [default=None]: the name of the PySCeS input file dir [default=pysces.model_dir]: the optional directory where the PSC file can be found """ if dir != None: if os.path.isdir(dir): pass else: chkmdir() dir = MODEL_DIR else: chkmdir() dir = MODEL_DIR mfgo = 0 if File == None: mfgo = 1 try: if File != None: File = chkpsc(File) if not os.path.exists(os.path.join(dir,File)): mfgo = 1 except: mfgo = 1 while mfgo == 1: print 'Models available in your model_dir: \n************' cntr = 0 namelen = 0 while len(os.listdir(dir)) == 0: print 'No models available in model directory, please set using pys_usercfg.ini or call\ with pysces.model(\'file.psc\',dir=\'path\\to\\models\')' dir = raw_input('\nPlease enter full path to models <CTRL+C> exits: ') dirList = os.listdir(dir) for x in range(len(dirList)-1,-1,-1): if dirList[x][-4:] != '.psc': a = dirList.pop(x) for x in dirList: if len(x) > namelen: namelen = len(x) for x in dirList: if cntr < 2: print x + ' '*(namelen-len(x)), cntr += 1 else: print x cntr = 0 print '\n************\n' print '\nYou need to specify a valid model file ...\n' File = raw_input('\nPlease enter filename: ') try: File = chkpsc(File) if os.path.exists(os.path.join(dir,File)): mfgo = 0 except: mfgo = 1 print 'Using model directory: ' + dir try: if os.path.exists(os.path.join(dir,File)): print os.path.join(dir,File) + ' loading .....', self.ModelDir = dir self.ModelFile = File else: print os.path.join(dir,File) + ' does not exist' print 'Please set with ModelDir and ModelFile .....', self.ModelFile = 'None' self.ModelDir = dir except: print os.path.join(dir,File) + ' does not exist please re-instantiate model .....', self.__settings__['display_debug'] = 0 self.ModelOutput = OUTPUT_DIR ## # Initialise elementary modes ## if os.sys.platform == 'win32': ## self.eModeExe_int = os.path.join(self.__metatool,'meta43_int.exe') ## self.eModeExe_dbl = os.path.join(self.__metatool,'meta43_double.exe') ## else: ## self.eModeExe_int = os.path.join(self.__metatool,'meta43_int') ## self.eModeExe_dbl = os.path.join(self.__metatool,'meta43_double') ## print 'Done.' # Initialize serial directory self.__settings__['serial_dir'] = os.path.join(self.ModelOutput,'pscdat') # Initialize stoichiometric precision self.__settings__['stoichiometric_analysis_fp_zero'] = mach_spec.eps*2.0e4 self.__settings__['stoichiometric_analysis_lu_precision'] = self.__settings__['stoichiometric_analysis_fp_zero'] self.__settings__['stoichiometric_analysis_gj_precision'] = self.__settings__['stoichiometric_analysis_lu_precision']*10.0 #def __LoadStomPyInterface__(self): #""" #Load the StomPy Stochastic simulation interface #""" #if _HAVE_STOMPY and self.__STOMPY__ == None: #self.__STOMPY__ = StomPyInterface(MODEL_DIR, OUTPUT_DIR) #print 'PySCeS/StomPy interface active'
def __ParsePiecewiseFunctions__(self, piecewises): """ THIS IS HIGHLY EXPERIMENTAL! Takes the a piecewises dictionary and creates a piecewise object while substituting the object name in the formula string. - *piecewises* a piecewise dictionary created by the InfixParser """ if piecewises != None and len(piecewises.keys()) > 0: self.__HAS_PIECEWISE__ = True for p in piecewises: print 'Info: adding piecewise object: %s' % p #if len(piecewises[p].keys()) == 2: # piecewises[p][0].reverse() # only for libsbml generated infix self.__piecewises__.update({p : piecewises[p]}) P = PieceWise(piecewises[p], self) P.setName(p) self.piecewise_functions.append(P) setattr(self, p, P)
[docs] def InitialiseInputFile(self): """ InitialiseInputFile() Parse the input file associated with the PySCeS model instance and assign the basic model attributes Arguments: None """ self.__parseOK = 1 # check that model has parsed ok? try: if os.path.exists(os.path.join(self.ModelDir,self.ModelFile)): pass else: print '\nInvalid self.ModelFile: ' + os.path.join(self.ModelDir,self.ModelFile) except: print 'WARNING: Problem verifying: ' + os.path.join(self.ModelDir,self.ModelFile) if self.ModelFile[-4:] == '.psc': pass else: print 'Assuming extension is .psc' self.ModelFile += '.psc' print '\nParsing file: %s' % os.path.join(self.ModelDir, self.ModelFile) pscParser.ParsePSC(self.ModelFile,self.ModelDir,self.ModelOutput) print ' ' badlist = pscParser.KeywordCheck(pscParser.ReactionIDs) badlist = pscParser.KeywordCheck(pscParser.Inits,badlist) if len(badlist) != 0: print '\n******************************\nPSC input file contains PySCeS keywords please rename them and reload:' for item in badlist: print ' --> ' + item print '******************************\n' self.__parseOK = 0 #assert len(badlist) != 0, 'Keyword error, please check input file' if self.__parseOK: # brett 2008 InfixParser.__pwcntr__ = 0 self.__nDict__ = pscParser.nDict.copy() self.__sDict__ = pscParser.sDict.copy() self.__pDict__ = pscParser.pDict.copy() self.__uDict__ = pscParser.uDict.copy() # model attributes are now initialised here brett2008 self.__InitDict__ = {} # set parameters and add to __InitDict__ for p in self.__pDict__.keys(): setattr(self, self.__pDict__[p]['name'], self.__pDict__[p]['initial']) self.__InitDict__.update({self.__pDict__[p]['name'] : self.__pDict__[p]['initial']}) # set species and add to __InitDict__ and set mod.Xi_init for s in self.__sDict__.keys(): setattr(self, self.__sDict__[s]['name'], self.__sDict__[s]['initial']) if not self.__sDict__[s]['fixed']: setattr(self, self.__sDict__[s]['name']+'_init', self.__sDict__[s]['initial']) self.__InitDict__.update({self.__sDict__[s]['name'] : self.__sDict__[s]['initial']}) # setup keywords self.__KeyWords__ = pscParser.KeyWords.copy() if self.__KeyWords__['Modelname'] == None: self.__KeyWords__['Modelname'] = self.ModelFile.replace('.psc','') if self.__KeyWords__['Description'] == None: self.__KeyWords__['Description'] = self.ModelFile.replace('.psc','') # if SpeciesTypes undefined assume [] if self.__KeyWords__['Species_In_Conc'] == None: self.__KeyWords__['Species_In_Conc'] = True # if OutputType is undefined assume it is the same as SpeciesType if self.__KeyWords__['Output_In_Conc'] == None: if self.__KeyWords__['Species_In_Conc']: self.__KeyWords__['Output_In_Conc'] = True else: self.__KeyWords__['Output_In_Conc'] = False if self.__KeyWords__['ModelType'] == None: self.__KeyWords__['ModelType'] = ['Deterministic'] else: self.__KeyWords__['ModelType'] = tuple([t.strip() for t in self.__KeyWords__['ModelType'].split(',')]) # we now check for modeltype, if it specified as stochastic check if stochpy is available #if self.__KeyWords__['ModelType'] == ['Stochastic']: #if _HAVE_STOMPY: #print 'INFO: This model suggests that it requires discrete simulation and StochPy is installed ... mod.doStochSim() and mod.doStochSimPlot() are available for use.' #else: #print 'INFO: This model suggests that it requires stochastic simulation and StochPy (stompy.sf.net) is not installed ... PySCeS will treat this model as continuous.' # set the species type in sDict according to 'Species_In_Conc' for s in self.__sDict__.keys(): if not self.__KeyWords__['Species_In_Conc']: self.__sDict__[s]['isamount'] = True else: self.__sDict__[s]['isamount'] = False # setup compartments self.__compartments__ = pscParser.compartments.copy() if len(self.__compartments__.keys()) > 0: self.__HAS_COMPARTMENTS__ = True else: self.__HAS_COMPARTMENTS__ = False # no (self.) self.__fixed_species__ = copy.copy(pscParser.fixed_species) self.__species__ = copy.copy(pscParser.species) self.__parameters__ = copy.copy(pscParser.parameters) self.__reactions__ = copy.copy(pscParser.reactions) self.__modifiers__ = copy.copy(pscParser.modifiers) # Initialize exposed stuff self.fixed_species = tuple(pscParser.fixed_species) self.species = tuple(pscParser.species) self.parameters = tuple(pscParser.parameters) self.reactions = tuple(pscParser.reactions) self.modifiers = tuple(pscParser.modifiers) # Add input file defined fuctions - brett 200500621 # TODO deprecated # self._Function_time = copy.copy(pscParser.TimeFunc) self._Function_user = copy.copy(pscParser.UserFunc) self._Function_init = pscParser.InitFunc self.__functions__ = pscParser.Functions.copy() self.__rules__ = pscParser.AssignmentRules.copy() self.__InitFuncs__ = pscParser.ModelInit.copy() self.__userfuncs__ = pscParser.UserFuncs.copy() self.__eDict__ = pscParser.Events.copy() self.__cbm_fluxbounds__ = pscParser.cbm_FluxBounds.copy() self.__cbm_objfuncs__ = pscParser.cbm_ObjectiveFunctions.copy() self.__cbm_userfluxconstraints__ = pscParser.cbm_UserFluxConstraints.copy() ## if pscParser.ModelUsesNumpyFuncs: ## print 'Numpy functions detected in kinetic laws.\n' else: print '\nERROR: model parsing error, please check input file.\n' # added in a check for model correctness and human error reporting (1=ok, 0=error) if len(pscParser.SymbolErrors) != 0: print '\nUndefined symbols:\n%s' % self.SymbolErrors if not pscParser.ParseOK: print '\n\n*****\nModel parsing errors detected in input file '+ self.ModelFile +'\n*****' print '\nInput file errors' for error in pscParser.LexErrors: print error ## try: ## print error[0] + 'in line:\t' + str(error[1]) + ' ('+ error[2][:20] +' ...)' ## except IndexError: ## print 'Illegal character:', error.__repr__(), error.__str__() print '\nParser errors' for error in pscParser.ParseErrors: print error ## try: ## print error[0] + '- ' + error[2][:20] ## except IndexError: ## print 'Illegal character:', error assert pscParser.ParseOK == 1, 'Input File Error'
[docs] def InitialiseRules(self): # we need to detect different types of rules etc #defmod self.__HAS_FORCED_FUNCS__ = False self.__HAS_RATE_RULES__ = False rate_rules = {} assignment_rules = {} for ar in self.__rules__: if self.__rules__[ar]['type'] == 'assignment': self.__HAS_FORCED_FUNCS__ = True assignment_rules.update({ar : self.__rules__[ar]}) elif self.__rules__[ar]['type'] == 'rate': self.__HAS_RATE_RULES__ = True rate_rules.update({ar : self.__rules__[ar]}) # THE NEW WAY (adds formula parsing for numpy functions etc) InfixParser.setNameStr('self.', '') self._NewRuleXCode = {} if self.__HAS_FORCED_FUNCS__ and not self.__settings__['mode_substitute_assignment_rules']: code_string = '' all_names = [] for ar in assignment_rules: name = assignment_rules[ar]['name'] InfixParser.setNameStr('self.', '') InfixParser.parse(assignment_rules[ar]['formula']) assignment_rules[ar]['symbols'] = InfixParser.names formula = InfixParser.output # this code has to stay together # for pw in InfixParser.piecewises: formula = formula.replace('self.%s' % pw, 'self.%s()' % pw) self.__ParsePiecewiseFunctions__(InfixParser.piecewises) # this code has to stay together # assignment_rules[ar]['code_string'] = formula all_names += InfixParser.names os.chdir(self.ModelOutput) keep = [] rules = assignment_rules.keys() dep = rules while len(dep) > 0: dep = [] indep = [] for ar in rules: if ar in all_names: indep.append(ar) else: dep.append(ar) keep += dep rules = indep for ar in indep+keep: evalCode = 'self.%s = %s\n' % (assignment_rules[ar]['name'],\ assignment_rules[ar]['code_string']) self._NewRuleXCode.update({assignment_rules[ar]['name'] : evalCode}) code_string += evalCode print '\nAssignment rule(s) detected.' self._Function_forced = code_string elif self.__HAS_FORCED_FUNCS__ and self.__settings__['mode_substitute_assignment_rules']: # here we substitute nested assignment rules InfixParser.setNameStr('self.', '') symbR = {} for ass in assignment_rules: InfixParser.setNameStr('self.', '') InfixParser.parse(assignment_rules[ass]['formula']) formula = InfixParser.output # this code has to stay together # for pw in InfixParser.piecewises: formula = formula.replace('self.%s' % pw, 'self.%s()' % pw) self.__ParsePiecewiseFunctions__(InfixParser.piecewises) # this code has to stay together # symbR.update({assignment_rules[ass]['name'] : formula}) for ass in assignment_rules: InfixParser.setNameStr('self.', '') InfixParser.FunctionReplacements = symbR InfixParser.parse(assignment_rules[ass]['formula']) assignment_rules[ass]['code_string'] = InfixParser.output self._Function_forced = 'pass\n' print 'Assignment rule(s) detected and substituted.' else: self._Function_forced = 'pass\n' self.__CODE_forced = compile(self._Function_forced,'AssignRules','exec') # tested in InitialiseRuleChecks() # defmod self.__rate_rules__ = [] self.__rrule__ = None rr_code_block = '' rr_map_block = '' self.__CODE_raterule = None self._NewRateRuleXCode = {} if self.__HAS_RATE_RULES__: # create a rr vector self.__rrule__ = numpy.ones(len(rate_rules.keys()),'d') # brett2008 debug stuff doesn't do any harm cntr = 0 rr_keys = rate_rules.keys() rr_keys.sort() for ar in rr_keys: name = rate_rules[ar]['name'] InfixParser.setNameStr('self.', '') InfixParser.parse(rate_rules[ar]['formula']) formula = InfixParser.output rate_rules[ar]['symbols'] = InfixParser.names # this code has to stay together # for pw in InfixParser.piecewises: formula = formula.replace('self.%s' % pw, 'self.%s()' % pw) self.__ParsePiecewiseFunctions__(InfixParser.piecewises) # this code has to stay together # rate_rules[ar]['code_string'] = formula self.__rate_rules__.append(name) rr_code_block += 'self.__rrule__[%s] = %s\n' % (cntr, formula) ## rr_code_block += 'self.%s = self.__rrule__[%s]\n' % (name, cntr) rr_map_block += 'self.%s = self.__rrule__[%s]\n' % (name, cntr) cntr += 1 # create mod.<rule name>_init attributes setattr(self, '%s_init' % name, getattr(self, name)) os.chdir(self.ModelOutput) print 'Rate rule(s) detected.' else: rr_code_block = 'pass\n' rr_map_block = 'pass\n' #TODO consider putting this in self.__HAS_RATE_RULES__ self.__CODE_raterule = compile(rr_code_block,'RateRules','exec') self.__CODE_raterule_map = compile(rr_map_block,'RateRuleMap','exec') del rate_rules, assignment_rules
[docs] def InitialiseRuleChecks(self): try: exec(self.__CODE_raterule) except ZeroDivisionError: print 'WARNING: Assignment RateRule ZeroDivision on initialisation (continuing)' except Exception, ex: print 'WARNING: RateRule initialisation error\n', ex zeroDivErr = [] for k in self._NewRuleXCode: try: exec(compile(self._NewRuleXCode[k],'NewRuleXCode','exec')) except Exception, ex: zeroDivErr.append(k) exit = len(self._NewRuleXCode.keys()) while len(zeroDivErr) > 0 and exit > 0: zeroDivErr2 = [] for kk in zeroDivErr: try: exec(compile(self._NewRuleXCode[kk],'NewRuleXCode','exec')) if self.__HAS_RATE_RULES__: exec(self.__CODE_raterule) exec(self.__CODE_raterule_map) except Exception, ex: zeroDivErr2.append(kk) exit -= 1 zeroDivErr = zeroDivErr2 if exit < 0: print 'WARNING: ZeroDivision elimination failed' try: exec(self.__CODE_forced) #print 'done.' except ZeroDivisionError: print 'WARNING: Assignment rule ZeroDivision on intialisation' except Exception, ex: print 'WARNING: Assignment rule error (disabling all rules)\n', ex self.__CODE_forced = compile('pass\n','AssignRules','exec')
[docs] def Stoichiometry_Init(self,nmatrix,load=0): """ Stoichiometry_Init(nmatrix,load=0) Initialize the model stoichiometry. Given a stoichiometric matrix N, this method will return an instantiated PyscesStoich instance and status flag. Alternatively, if load is enabled, PySCeS will attempt to load a previously saved stoichiometric analysis (saved with Stoichiometry_Save_Serial) and test it's correctness. The status flag indicates 0 = reanalyse stoichiometry or 1 = complete structural analysis preloaded. Arguments: nmatrix: The input stoichiometric matrix, N load [default=0]: try to load a saved stoichiometry (1) """ if load: print 'Loading saved stoichiometry ...' try: stc = self.Stoichiometry_Load_Serial() go = 1 except Exception, slx: print slx go = 0 row,col = nmatrix.shape if go: badList = [] for x in range(row): if stc.__species__[x] != self.__species__[x]: badList.append((stc.__species__[x],self.__species__[x])) go = 0 for y in range(col): if stc.__reactions__[y] != self.__reactions__[y]: badList.append((stc.__reactions__[y],self.__reactions__[y])) go = 0 if abs(nmatrix[x,y] - stc.nmatrix[x,y]) > stc.stoichiometric_analysis_fp_zero: badList.append(((x,y),nmatrix[x,y],stc.nmatrix[x,y])) go = 0 if not go: ## print 'Stoichiometry mismatch' ## for x in badList: ## print x, print '\nProblem loading stoichiometry, reanalysing ...' stc = PyscesStoich.Stoich(nmatrix) status = 0 else: print 'Stoichiometry verified ... we have liftoff' status = 1 else: #print 'Instantiating new stoichiometry ...' stc = PyscesStoich.Stoich(nmatrix) status = 0 return stc,status
[docs] def Stoichiometry_Save_Serial(self): """Serialize and save a Stoichiometric instance to binary pickle"""""" Stoichiometry_Save_Serial() Serilaise and save the current model stoichiometry to a file with name <model>_stoichiometry.pscdat in the mod.__settings__['serial_dir'] directory (default: mod.model_output/pscdat) Arguments: None """ # new plan, I introduce species and reaction arrays into the stoich instance for verification purposes # brett - 20050831 self.__structural__.__species__ = copy.copy(self.__species__) self.__structural__.__reactions__ = copy.copy(self.__reactions__) self.SerialEncode(self.STOICH,self.ModelFile[:-4]+'_stoichiometry')
[docs] def Stoichiometry_Load_Serial(self): """ Stoichiometry_Load_Serial() Load a saved stoichiometry saved with mod.Stoichiometry_Save_Serial() and return a stoichiometry instance. Arguments: None """ stc = self.SerialDecode(self.ModelFile[:-4]+'_stoichiometry') return stc # new powerslave method - brett 20050825
[docs] def Stoichiometry_Analyse(self,override=0,load=0): """ Stoichiometry_Analyse(override=0,load=0) Perform a structural analyses. The default behaviour is to construct and analyse the model from the parsed model information. Overriding this behaviour analyses the stoichiometry based on the current stoichiometric matrix. If load is specified PySCeS tries to load a saved stoichiometry, otherwise the stoichiometric analysis is run. The results of the analysis are checked for floating point error and nullspace rank consistancy. Arguments: override [default=0]: override stoichiometric analysis intialisation from parsed data load [default=0]: load a presaved stoichiometry """ if not override: self.__nmatrix__ = self.__initmodel__() #Creates the model N #print '\nintializing N\n' else: print '\nStoichiometric override active\n' assert len(self.__nmatrix__) > 0, '\nUnable to generate Stoichiometric Matrix! model has:\n%s reactions\n%s species\nwhat did you have in mind?\n' % (len(self.__reactions__), len(self.__species__)) self.__Nshape__ = self.__nmatrix__.shape #Get the shape of N ## self.__Vtemp__ = numpy.zeros((self.__Nshape__[1])) # going going .... # get stoich instance and whether it was analysed or loaded - brett 20050830 self.__structural__, stc_load = self.Stoichiometry_Init(self.__nmatrix__,load=load) # if not loaded analyze - brett 20050830 if not stc_load: # technically this means we can define this on the fly - brett #20051013 self.__structural__.stoichiometric_analysis_fp_zero = self.__settings__['stoichiometric_analysis_fp_zero'] self.__structural__.stoichiometric_analysis_lu_precision = self.__settings__['stoichiometric_analysis_lu_precision'] self.__structural__.stoichiometric_analysis_gj_precision = self.__settings__['stoichiometric_analysis_gj_precision'] self.__structural__.AnalyseL() #Get all L related stuff self.__structural__.AnalyseK() #Get all K related stuff #test matrix values against __settings__['stoichiometric_analysis_lu_precision'] lsmall,lbig = self.__structural__.MatrixValueCompare(self.__structural__.lzeromatrix) ksmall,kbig = self.__structural__.MatrixValueCompare(self.__structural__.kzeromatrix) SmallValueError = 0 if abs(lsmall) < self.__structural__.stoichiometric_analysis_lu_precision*10.0: print '\nWARNING: values in L0matrix are close to stoichiometric precision!' print 'Stoichiometric LU precision:', self.__structural__.stoichiometric_analysis_lu_precision print 'L0 smallest abs(value)', abs(lsmall) print 'Machine precision:', mach_spec.eps SmallValueError = 1 if abs(ksmall) < self.__structural__.stoichiometric_analysis_lu_precision*10.0: print '\nWARNING: values in K0matrix are close to stoichiometric precision!' print 'Stoichiometric precision:', self.__structural__.stoichiometric_analysis_lu_precision print 'K0 smallest abs(value)', abs(ksmall) print 'Machine precision:', mach_spec.eps SmallValueError = 1 if SmallValueError: raw_input('\nStructural Analysis results may not be reliable!!!.\n\nTry change <mod>.__settings__["stoichiometric_analysis_lu_precision"] (see reference manual for details)\n\n\t press any key to continue: ') # cross check that rank is consistant between K0 and L0 if self.__structural__.kzeromatrix.shape[0] != self.__structural__.lzeromatrix.shape[1]: print '\nWARNING: the rank calculated by the Kand L analysis methods are not the same!' print '\tK analysis calculates the rank as: ' + `self.__structural__.kzeromatrix.shape[0]` print '\tL analysis calculates the rank as: ' + `self.__structural__.lzeromatrix.shape[1]` print 'This is not good! Structural Analysis results are not reliable!!!\n' assert self.__structural__.kzeromatrix.shape[0] == self.__structural__.lzeromatrix.shape[1], '\nStructuralAnalysis Error: rank mismatch' self.__HAS_FLUX_CONSERVATION__ = self.__structural__.info_flux_conserve self.__HAS_MOIETY_CONSERVATION__ = self.__structural__.info_moiety_conserve if self.__settings__['enable_deprecated_attr']: ## self.__nmatrix__ = copy.copy(self.nmatrix) self.nmatrix = self.__nmatrix__ # done with caution brett2008 self.nmatrix_row = self.__structural__.nmatrix_row self.nmatrix_col = self.__structural__.nmatrix_col self.kmatrix = self.__structural__.kmatrix self.kmatrix_row = self.__structural__.kmatrix_row self.kmatrix_col = self.__structural__.kmatrix_col self.kzeromatrix = self.__structural__.kzeromatrix self.kzeromatrix_row = self.__structural__.kzeromatrix_row self.kzeromatrix_col = self.__structural__.kzeromatrix_col self.lmatrix = self.__structural__.lmatrix self.lmatrix_row = self.__structural__.lmatrix_row self.lmatrix_col = self.__structural__.lmatrix_col self.lzeromatrix = self.__structural__.lzeromatrix self.lzeromatrix_row = self.__structural__.lzeromatrix_row self.lzeromatrix_col = self.__structural__.lzeromatrix_col self.conservation_matrix = self.__structural__.conservation_matrix self.conservation_matrix_row = self.__structural__.conservation_matrix_row self.conservation_matrix_col = self.__structural__.conservation_matrix_col self.nrmatrix = self.__structural__.nrmatrix self.nrmatrix_row = self.__structural__.nrmatrix_row self.nrmatrix_col = self.__structural__.nrmatrix_col self.__kmatrix__ = copy.copy(self.kmatrix) self.__kzeromatrix__ = copy.copy(self.kzeromatrix) self.__lmatrix__ = copy.copy(self.lmatrix) self.__lzeromatrix__ = copy.copy(self.lzeromatrix) self.__nrmatrix__ = copy.copy(self.nrmatrix) # switch that is set if the stoichiometric analysis is up to date self.__structural__.species = self.species self.__structural__.reactions = self.reactions self.Nmatrix = PyscesStoich.StructMatrix(self.__structural__.nmatrix, self.__structural__.nmatrix_row, self.__structural__.nmatrix_col) self.Nmatrix.setRow(self.species) self.Nmatrix.setCol(self.reactions) self.Nrmatrix = PyscesStoich.StructMatrix(self.__structural__.nrmatrix, self.__structural__.nrmatrix_row, self.__structural__.nrmatrix_col) self.Nrmatrix.setRow(self.species) self.Nrmatrix.setCol(self.reactions) self.Kmatrix = PyscesStoich.StructMatrix(self.__structural__.kmatrix, self.__structural__.kmatrix_row, self.__structural__.kmatrix_col) self.Kmatrix.setRow(self.reactions) self.Kmatrix.setCol(self.reactions) self.K0matrix = PyscesStoich.StructMatrix(self.__structural__.kzeromatrix, self.__structural__.kzeromatrix_row, self.__structural__.kzeromatrix_col) self.K0matrix.setRow(self.reactions) self.K0matrix.setCol(self.reactions) self.Lmatrix = PyscesStoich.StructMatrix(self.__structural__.lmatrix, self.__structural__.lmatrix_row, self.__structural__.lmatrix_col) self.Lmatrix.setRow(self.species) self.Lmatrix.setCol(self.species) self.L0matrix = PyscesStoich.StructMatrix(self.__structural__.lzeromatrix, self.__structural__.lzeromatrix_row, self.__structural__.lzeromatrix_col) self.L0matrix.setRow(self.species) self.L0matrix.setCol(self.species) if self.__structural__.info_moiety_conserve: self.Consmatrix = PyscesStoich.StructMatrix(self.__structural__.conservation_matrix, self.__structural__.conservation_matrix_row, self.__structural__.conservation_matrix_col) self.Consmatrix.setRow(self.species) self.Consmatrix.setCol(self.species) else: self.Consmatrix = None self.__StoichOK = 1 print ' '
[docs] def Stoichiometry_ReAnalyse(self): """ Stoichiometry_ReAnalyse() Reanalyse the stoichiometry using the current N matrix ie override=1 (for use with mod.Stoich_matrix_SetValue) Arguments: None """ self.Stoichiometry_Analyse(override=1) self.InitialiseConservationArrays()
[docs] def Stoich_nmatrix_SetValue(self,species,reaction,value): """ Stoich_nmatrix_SetValue(species,reaction,value) Change a stoichiometric coefficient's value in the N matrix. Only a coefficients magnitude may be set, in other words a a coefficient's value must remain negative, positive or zero. After changing a coefficient it is necessary to Reanalyse the stoichiometry. Arguments: species: species name (s0) reaction: reaction name (R4) value: new coefficient value """ index = self.__Stoich_nmatrix_FindIndex__(species,reaction) if self.__Stoich_nmatrix_CheckValue__(index,value): self.__Stoich_nmatrix_UpdateValue__(index,value) self.__StoichOK = 0
def __Stoich_nmatrix_FindIndex__(self,species,reaction): """ __Stoich_nmatrix_FindIndex__(species,reaction) Return the N matrix co-ordinates of the coefficient referenced by species and reaction name. Arguments: species: species name reaction: reaction name """ rval = None cval = None for x in enumerate(self.species): if species == x[1]: rval = x[0] for y in enumerate(self.reactions): if reaction == y[1]: cval = y[0] return (rval,cval) def __Stoich_nmatrix_UpdateValue__(self,(x,y),val): """ __Stoich_nmatrix_UpdateValue__((x,y), val) Update N matrix co-ordinates (x,y) with val Arguments: (x,y): row, column coordinates val: value """ self.nmatrix[x,y] = val self.__nmatrix__[x,y] = val self.Nmatrix.array[x,y] = val def __Stoich_nmatrix_CheckValue__(self,(x,y),val): """ __Stoich_nmatrix_CheckValue__((x,y),val) Check validity of new coefficient value against existing N matrix coefficient (x,y) for existance and/or sign. Returns 1 (true) or 0. Arguments: (x,y): N matrix coordinates val: new value """ go = 1 if x == None or y == None: print '\nSpecies (nmatrix) index =', x print 'Reaction (nmatrix) index =', y print '\nI\'m confused, perhaps you entered an incorrect species or reaction name' print 'or they are in the wrong order?' go = 0 elif abs(self.__nmatrix__[x,y]) == 0.0: if val != 0.0: go = 0 print '\nZero coefficient violation' print ' nmatrix['+`x`+','+`y`+'] can only be = 0.0 (input '+str(val)+')' elif self.__nmatrix__[x,y] > 0.0: if val <= 0.0: go = 0 print '\nPositive coefficient violation' print ' nmatrix['+`x`+','+`y`+'] can only be > 0.0 (input '+str(val)+')' elif self.__nmatrix__[x,y] < 0.0: if val >= 0.0: go = 0 print '\nNegative coefficient violation' print ' nmatrix['+`x`+','+`y`+'] can only be < 0.0 (input '+str(val)+')' if go: return 1 else: return 0
[docs] def SerialEncode(self,data,filename): """ SerialEncode(data,filename) Serialise and save a Python object using a binary pickle to file. The serialised object is saved as <filename>.pscdat in the directory defined by mod.model_serial. Arguments: data: pickleable Python object filename: the ouput filename """ filename = str(filename)+'.pscdat' if os.path.exists(self.__settings__['serial_dir']): pass else: os.mkdir(self.__settings__['serial_dir']) print '\ncPickle data stored in: ' + self.__settings__['serial_dir'] File = open(os.path.join(self.__settings__['serial_dir'], filename),'wb') try: cPickle.dump(data,File,-1) print 'Serialization complete' except Exception, E: print 'Serialize error:\n',E File.flush() File.close() del data
[docs] def SerialDecode(self,filename): """ SerialDecode(filename) Decode and return a serialised object saved with SerialEncode. Arguments: filename: the filename (.pscdat is assumed) """ filename = str(filename)+'.pscdat' if not os.path.exists(os.path.join(self.__settings__['serial_dir'],filename)): raise RuntimeError, 'Serialized data '+os.path.join(self.__settings__['serial_dir'],filename)+' does not exist' data = None try: File = open(os.path.join(self.__settings__['serial_dir'], filename),'rb') data = cPickle.load(File) File.close() print 'Serial decoding complete' except Exception, E: print 'Serial decode error:\n',E return data
[docs] def InitialiseCompartments(self): # the name should say it all self.__settings__['compartment_fudge_factor'] = 1.0 startFudging = 1.0e-8 I_AM_FUDGING = False if self.__HAS_COMPARTMENTS__: tmp = min([abs(self.__compartments__[c]['size']) for c in self.__compartments__.keys()]) if tmp < startFudging: ## self.__settings__['compartment_fudge_factor'] = 10.0**round(numpy.log10(tmp)) self.__settings__['compartment_fudge_factor'] = tmp # sneaky b%*))_)%$^&*@rds ## print 'compartment_fudge_factor', self.__settings__['compartment_fudge_factor'] for c in self.__compartments__.keys(): if self.__settings__['compartment_fudge_factor'] < startFudging: newsize = self.__compartments__[c]['size']/self.__settings__['compartment_fudge_factor'] print 'INFO: Rescaling compartment with size %s to %s' % (self.__compartments__[c]['size'], newsize) self.__compartments__[c]['size'] = newsize self.__compartments__[c].update({'scale' : self.__settings__['compartment_fudge_factor']}) I_AM_FUDGING = True setattr(self, self.__compartments__[c]['name'], self.__compartments__[c]['size']) setattr(self, '%s_init' % self.__compartments__[c]['name'], self.__compartments__[c]['size']) if self.__HAS_COMPARTMENTS__: for sp in self.__sDict__.keys(): if self.__sDict__[sp]['compartment'] == None and len(self.__compartments__.keys()) == 1: self.__sDict__[sp]['compartment'] = self.__compartments__[self.__compartments__.keys()[0]]['name'] print 'COMPARTMENT WARNING: this model has a compartment defined %s but \"%s\" is not in it ... I\'ll try it for now' %\ ([self.__compartments__[c]['name'] for c in self.__compartments__.keys()], sp) ## print self.__sDict__[sp] elif self.__sDict__[sp]['compartment'] == None and len(self.__compartments__.keys()) > 1: assert self.__sDict__[sp]['compartment'] != None,\ '\nCOMPARTMENT ERROR: this model has multiple compartments defined %s but \"%s\" is not in one!' %\ ([self.__compartments__[c]['name'] for c in self.__compartments__.keys()], sp) # brett 2008 fudge this! if not self.__KeyWords__['Species_In_Conc'] and I_AM_FUDGING: self.__sDict__[sp]['initial'] = self.__sDict__[sp]['initial']/self.__settings__['compartment_fudge_factor'] setattr(self, sp, self.__sDict__[sp]['initial']) setattr(self, '%s_init' % sp, self.__sDict__[sp]['initial']) print 'INFO: Rescaling species (%s) to size %s.' % (sp, self.__sDict__[sp]['initial']) self.__CsizeAllIdx__ = [] self.__null_compartment__ = 1.0 uses_null_compartments = False for s in self.__species__: if self.__HAS_COMPARTMENTS__: self.__CsizeAllIdx__.append(self.__sDict__[s]['compartment']) else: self.__CsizeAllIdx__.append('__null_compartment__') uses_null_compartments = True if uses_null_compartments: setattr(self, '__null_compartment___init', 1.0) ## self.__CsizeFSIdx__ = [] ## for s in self.__fixed_species__: ## if self.__HAS_COMPARTMENTS__: ## self.__CsizeFSIdx__.append(self.__sDict__[s]['compartment']) ## else: ## self.__CsizeFSIdx__.append('__null_compartment__') # Init compartment divisions vector - brett 2008 ## self.__CsizeAll__ = numpy.ones(len(self.__CsizeAllIdx__), 'd') if self.__HAS_COMPARTMENTS__: self.__CsizeAll__ = numpy.array([self.__compartments__[c]['size'] for c in self.__CsizeAllIdx__], 'd') else: self.__CsizeAll__ = numpy.ones(len(self.__CsizeAllIdx__), 'd') ## self.__CsizeFS__ = numpy.array([self.__compartments__[c]['size'] for c in self.__CsizeFSIdx__], 'd') # add new function types to model
[docs] def InitialiseFunctions(self): for func in self.__functions__: F = Function(func, self) for arg in self.__functions__[func]['args']: F.setArg(arg.strip()) F.addFormula(self.__functions__[func]['formula']) setattr(self, func, F) os.chdir(self.ModelOutput) for func in self.__functions__: fobj = getattr(self, func) for f in fobj.functions: if f in self.__functions__: setattr(fobj, f, getattr(self, f)) # add event types to model
[docs] def InitialiseEvents(self): self.__events__ = [] # for each event for e in self.__eDict__: ev = Event(e, self) ev._time_symbol = self.__eDict__[e]['tsymb'] ev.setTrigger(self.__eDict__[e]['trigger'], self.__eDict__[e]['delay']) # for each assignment for ass in self.__eDict__[e]['assignments']: ev.setAssignment(ass, self.__eDict__[e]['assignments'][ass]) self.__events__.append(ev) setattr(self, ev.name, ev) os.chdir(self.ModelOutput) if len(self.__events__) > 0: self.__HAS_EVENTS__ = True print 'Event(s) detected.' else: self.__HAS_EVENTS__ = False
[docs] def InitialiseModel(self): """ InitialiseModel() Initialise and set up dynamic model attributes and methods using the model defined in the associated PSC file Arguments: None """ #TODO killkillkill self.__inspec__ = numpy.zeros((len(self.__species__)),'d') self.__D_s_Order, DvarUpString = self.__initvar__() #Initialise s[] array self.__CDvarUpString__ = compile(DvarUpString,'DvarUpString','exec') self.state_species = None self.state_flux = None ## self.sim_res = None self.elas_var_u = None self.elas_var = None self.__vvec__ = numpy.zeros((self.__Nshape__[1]),'d') self.__tvec_a__ = None self.__tvec_c__ = None self.__CVODE_Vtemp = numpy.zeros((self.__Nshape__[1]),'d') # #debug # self.display_debugcount = 0 # self.cntr = 50 par_map2store, par_remap = self.__initfixed__() #Initialise fixed species #self.__par_map2storeC = par_map2store #self.__par_remapC = par_remap self.__par_map2storeC = compile(par_map2store,'par_map2store','exec') self.__par_remapC = compile(par_remap,'par_remap','exec') ## self.__xMake = compile(xString,'xString','exec') ## exec(self.__xMake) # initialise rate equations mapString,mapString_R,vString = self.__initREQ__() self.__mapFunc__ = compile(mapString,'mapString','exec') self.__mapFunc_R__ = compile(mapString_R,'mapString_R','exec') self.__vFunc__ = compile(vString,'vString','exec') ## exec(lambdaFuncsC) # Machine specific values (for IEEE compliant FP machines this is around 2.e-16) # FutureNote: investigate arbitrary precision FP in Python, probably GNU-GMP based - brett 20040326 self.__settings__['mach_floateps'] = mach_spec.eps # PySCeS mode switches self.__settings__['mode_number_format'] = '%2.4e' # this affects number output formatting in PySCeS) self.__settings__['mode_sim_init'] = 0 # 0:initval, 1:zeroval, 2:lastss 3:sim_res[-1] self.__settings__['mode_sim_max_iter'] = 3 # maximum number of auto-stepsize adjustments #TODO UPGRADE self.mode_state_init = 0 # 0:initval, 1:zeroval, 2:sim, 3:%state self.mode_solver = 'HYBRD' # 0:HYBRD, 1:FINTSLV, 2:NLEQ2 self.mode_solver_fallback = 1 # 0:Solver failure fails, # 1:solver failure falls back to NLEQ2 if present then FINTSLV (see next) self.STATE_extra_output = [] # extra data added to data_sstate object self.__settings__['mode_state_init3_factor'] = 0.1 # factor to multiply the previous ss used as initialiser self.__mode_state_init2_array__ = numpy.logspace(0,5,18) # the simulation array used to initialise the steady state self.__settings__['mode_state_mesg'] = 1 # happy exit message from State() self.__settings__['mode_state_nan_on_fail'] = False # give last best solutions or NaN as solution self.__settings__['solver_switch_warning'] = True # the irritating switching to message self.__settings__['mode_solver_fallback_integration'] = 1 # 0:no fallback to forward integration 1:fallback to integration # this is now initialised in the model parsing sectiomn ParseModel self.__settings__['mode_elas_deriv_order'] = 3 # the order of ScipyDerivative - 3 seems good for most situations self.__settings__['mode_mca_scaled'] = 1 # MCA scaling option 0:unscaled E+C+Eig 1:scaled E+C+Eig self.__settings__['mode_eigen_output'] = 0 # 0:normal, 1:extended eigen value + left/right vectors # under consideration self.__settings__['mode_suppress_info'] = 0 # 0:warnings displayed, 1:warnings suppressed # new compartment stuff (using dictionary directly) - brett 2008 ## self.Species_In_Conc = False # misc settings self.__settings__['write_array_header'] = 1 # write an id header before the array self.__settings__['write_array_spacer'] = 1 # write a empty line before the array self.__settings__['write_array_html_header'] = 1 #write html page header self.__settings__['write_array_html_footer'] = 1 #write html page footer self.__settings__['write_array_html_format'] = '%2.4f' #html number format self.__settings__['write_arr_lflush'] = 5 # lines to write before flushing to disk # Hybrd options (zero value means routine decides) self.__settings__['hybrd_xtol'] = 1.e-12 # relative error tolerance self.__settings__['hybrd_maxfev'] = 0 # Maximum number of calls, zero means then 100*(len(species)+1) self.__settings__['hybrd_epsfcn'] = copy.copy(self.__settings__['mach_floateps']) # A suitable step length for the forward-difference approximation of the Jacobian self.__settings__['hybrd_factor'] = 100 # A parameter determining the initial step bound in interval (0.1,100) self.__settings__['hybrd_mesg'] = 1 # print the exit status message # Finstslv options (forward integration solver) - brett (20030331) self.__settings__['fintslv_tol'] = 1.e-3 # max allowed deviation between max(sim_res) to be a steady state self.__settings__['fintslv_step'] = 5 # threshold number of steps where deviation < atol to be declared a steady state # a "saturation" type range - should be ok for most systems self.__fintslv_range__ = numpy.array([1,10,100,1000,5000,10000,50000,50100,50200,50300,50400,50500,50600,50700,50800,50850,50900,50950,51000],'d') #self.__fintslv_range__ = numpy.array([1,10,100,1000,5000,10000,50000,100000,500000,1000000,1000100, 1000200,1000300,1000400,1000500,1000600,1000700,1000800],'d') self.__settings__['fintslv_rmult'] = 1.0 # NLEQ2 options (optional non-linear solver) - brett (20030331) # IOPT self.__settings__['nleq2_advanced_mode'] = False # use advanced NLEQ2 features ... may speedup some parameter scans self.__settings__['nleq2_growth_factor'] = 10 # nitmax growth factor per nleq2 iteration self.__settings__['nleq2_iter'] = 3 # number of itertions to loop the solver through (2 should almost always be sufficient) self.__settings__['nleq2_iter_max'] = 10 # maximum numbe rof iterations in advanced mode self.__settings__['nleq2_rtol'] = 1.e-8 # automatically calculated by nleq12 self.__settings__['nleq2_jacgen'] = 2 # 2:numdiff, 3:numdiff+feedback self.__settings__['nleq2_iscaln'] = 0 # 0:xscal lower thresholdof scaling vector, 1:always scaling vector # Dangerous anything not zero seqfaults ## self.__settings__['nleq2_mprerr'] = 1 # 0:no output, 1:error, 2:+warning, 3:+info self.__settings__['nleq2_nonlin'] = 4 # 1:linear, 2:mildly non-lin, 3:highly non-lin, 4:extremely non-lin self.__settings__['nleq2_qrank1'] = 1 # 0:no Broyden approx. rank-1 updates, 1:Broyden approx. rank-1 updates self.__settings__['nleq2_qnscal'] = 0 # 0:Automatic row scaling is active, 1:inactive self.__settings__['nleq2_ibdamp'] = 0 # 0:auto damping strategy, 1:damping on, 2:damping off self.__settings__['nleq2_nitmax_growth_factor'] = 4 # on non-superlinear convergance nitmax *= this (ierr=5) # TODO: chekc this self.__settings__['nleq2_iormon'] = 2 # 0:default(2) 1:convergance not checked, 2:+'weak stop', 3:+'hard stop' criterion self.__settings__['nleq2_mesg'] = 1 # print the exit status message #TODO: self.nleq2_nitmax = 50 # optimized and self adjusting :-) # Other SteadyState options self.__settings__['small_concentration'] = 1.0e-6 # the initial values of 1:zeroval smaller than 1.0e-6 sometimes give problems ## self.__state_set_conserve__ = 1 self.__StateOK__ = True self.data_sstate = None self.data_sim = None # the new integration results data object self.data_stochsim = None # the new stochastic integration results data object self.__scan2d_results__ = None # yaaaay gnuplot is back self.__scan2d_pars__ = None # Simulate/lsoda options (zero value means routine decides) self.__settings__['lsoda_atol'] = 1.e-12 # The input parameters rtol and atol determine the error self.__settings__['lsoda_rtol'] = 1.e-7 # control performed by the solver. -- johann 20050217 changed from 1e-5 self.__settings__["lsoda_mxstep"] = 0 # maximum number (internally defined) steps allowed per point. 0: x <= 500 self.__settings__["lsoda_h0"] = 0.0 # the step size to be attempted on the first step. self.__settings__["lsoda_hmax"] = 0.0 # the maximum absolute step size allowed. self.__settings__["lsoda_hmin"] = 0.0 # the minimum absolute step size allowed. self.__settings__["lsoda_mxordn"] = 12 # maximum order to be allowed for the nonstiff (Adams) method. self.__settings__["lsoda_mxords"] = 5 # maximum order to be allowed for the stiff (BDF) method. self.__settings__["lsoda_mesg"] = 1 # print the exit status message # try to select the best integration algorithm self.mode_integrator = 'LSODA' # LSODA/CVODE set the intgration algorithm if self.__HAS_EVENTS__: if _HAVE_PYSUNDIALS: print '\nINFO: events detected and we have PySundials installed, switching to CVODE (mod.mode_integrator=\'CVODE\').' self.mode_integrator = 'CVODE' else: print '\nWARNING: PySCeS needs PySundials installed for event handling, PySCeS will continue with LSODA (NOTE: ALL EVENTS WILL BE IGNORED!).' print 'Windows users may install the "pysces_pysundials" package from pysces.sf.net' print 'For other OS\'s see pysundials.sf.net for installation details\n' self.__events__ = [] #if self.__HAS_RATE_RULES__ or self.__HAS_COMPARTMENTS__: if self.__HAS_RATE_RULES__: if _HAVE_PYSUNDIALS: print 'INFO: RateRules detected and PySundials installed, switching to CVODE (mod.mode_integrator=\'CVODE\').\n' self.mode_integrator = 'CVODE' else: print '\nWARNING: RateRules detected! PySCeS prefers CVODE but will continue with LSODA (NOTE: VARIABLE COMPARTMENTS ARE NOT SUPPORTED WITH LSODA!)' print 'Windows users may install the "pysces_pysundials" package from pysces.sf.net' print 'For other OS\'s see pysundials.sf.net for installation details\n' # pysundials self.mode_integrate_all_odes = False # only available with CVODE self.__settings__["cvode_abstol"] = 1.0e-15 # absolute tolerance ## self.__settings__["cvode_abstol_max"] = 1.0e-3 # not used anymore self.__settings__["cvode_abstol_factor"] = 1.0e-6 self.__settings__["cvode_reltol"] = 1.0e-9 # relative tolerance self.__settings__["cvode_auto_tol_adjust"] = True # auto reduce tolerances self.__settings__["cvode_mxstep"] = 1000 # max step default self.__settings__["cvode_stats"] = False # print some pretty stuff after a simulation if self.__HAS_PIECEWISE__ and self.__settings__["cvode_reltol"] <= 1.0e-9 : self.__settings__["cvode_reltol"] = 1.0e-6 print 'INFO: Piecewise functions detected increasing CVODE tolerance slightly (mod.__settings__[\"cvode_reltol\"] = 1.0e-9 ).' # Normal simulation options self.sim_start = 0.0 self.sim_end = 10.0 self.sim_points = 20.0 sim_steps = (self.sim_end - self.sim_start)/self.sim_points self.sim_time = numpy.arange(self.sim_start, self.sim_end + sim_steps, sim_steps) self.CVODE_extra_output = [] self.CVODE_xdata = None self.__SIMPLOT_OUT__ = [] # elasticity options self.__settings__["elas_evar_upsymb"] = 1 # attach elasticities self.__settings__["elas_epar_upsymb"] = 1 # attach parameter elasticities self.__settings__["elas_evar_remap"] = 1 # remap steady state values ... no if SimElas self.__settings__['mode_elas_deriv_factor'] = 0.0001 # used to determine a stepsize dx=So*factor self.__settings__['mode_elas_deriv_min'] = 1.0e-12 # minimum value dx is allowed to have self.__settings__['elas_zero_flux_fix'] = False # replaces zero fluxes with a very small number self.__settings__['elas_zero_conc_fix'] = False # replaces zero concentrations with a very small number self.__settings__['elas_scaling_div0_fix'] = False # if Infinite values are created when scaling set to zero # MCA options self.__settings__["mca_ccj_upsymb"] = 1 # attach the flux control coefficients self.__settings__["mca_ccs_upsymb"] = 1 # attach the concentration control coefficients self.__settings__["mca_ccall_fluxout"] = 1 # in .showCC() output flux cc's self.__settings__["mca_ccall_concout"] = 1 # in .showCC() output conc cc's self.__settings__["mca_ccall_altout"] = 0 # in .showCC() all CC's group by reaction # gone in 60 seconds brett2008 (Refactored to PyscesLink) ## # Elementary mode options ## self.emode_userout = 0 # write metatool output to file 0:no, 1:yes ## self.emode_intmode = 0 # 0:float metatool, 1:integer metatool ## self.emode_file = self.ModelFile[:-4] + '_emodes' # pitcon (pitcon 6.1) continuation options - brett 20040429 #my interface controls self.__settings__["pitcon_fix_small"] = 0 #in the REq evaluation values <1.e-15 = 1.e-15 #TODO: self.pitcon_par_space = numpy.logspace(-1,3,10) #parameter space that pitcon must search in self.pitcon_iter = 10 #number of iterations to search for every point in par_space self.__settings__["pitcon_allow_badstate"] = 0 #initialize with non steady-state values 0:no,1:yes self.__settings__["pitcon_flux_gen"] = 1 #generate fluxes as output self.__settings__["pitcon_filter_neg"] = 1 #drop pitcon results containing negative concentrations 0:no,1:yes self.__settings__["pitcon_filter_neg_res"] = 0 #drop output results containing negative concentrations 0:no,1:yes self.__settings__["pitcon_max_step"] = 30.0 #Maximum stepsize #TODO: self.pitcon_target_points = [] #list of calculated target points self.pitcon_limit_points = [] #list of calculated limit points self.pitcon_targ_val = 0.0 #Target value (Seek solution with iwork[4]=) # moved to InitialiseConservationArrays # self.__settings__["pitcon_max_step"] = 10*(len(self.__SI__)+1) #max corrector steps #pitcon integer options iwork in pitcon/dpcon61.f self.__settings__["pitcon_init_par"] = 1 #Use X(1) for initial parameter self.__settings__["pitcon_par_opt"] = 0 #Parameterization option 0:allows program self.__settings__["pitcon_jac_upd"] = 0 #Update jacobian every newton step self.__settings__["pitcon_targ_val_idx"] = 0 #Seek target values for X(n) self.__settings__["pitcon_limit_point_idx"] = 0 #Seek limit points in X(n) self.__settings__["pitcon_output_lvl"] = 0 #Control amount of output. self.__settings__["pitcon_jac_opt"] = 1 #Jacobian choice. 0:supply jacobian,1:use forward difference,2:central difference #pitcon float options rwork in pitcon/dpcon61.f self.__settings__["pitcon_abs_tol"] = 0.00001 #Absolute error tolerance self.__settings__["pitcon_rel_tol"] = 0.00001 #Relative error tolerance self.__settings__["pitcon_min_step"] = 0.01 #Minimum stepsize self.__settings__["pitcon_start_step"] = 0.3 #Starting stepsize self.__settings__["pitcon_start_dir"] = 1.0 #Starting direction +1.0/-1.0 self.__settings__["pitcon_max_grow"] = 3.0 #maximum growth factor # setup conservation matrices (L_switch is set by stoich to 1 if it "detects" conservation) self.InitialiseConservationArrays() # scan option - removed from initsimscan self.scan_in = '' self.scan_out = [] self._scan = None self.scan_res = None self.__settings__["scan1_mca_mode"] = 0 # should scan1 run mca analysis 0:no,1:elas,2:cc self.__settings__["scan1_dropbad"] = 0 # should scan1 drop invalid steady states (rare)? self.__settings__["scan1_nan_on_bad"] = True # invalid steady states are returned as NaN self.__settings__["scan1_mesg"] = True # print out progress messages for large (>20) scans self.__scan_errors_par__ = None # collect errors, parameters values self.__scan_errors_idx__ = None # collect errors, indexes
[docs] def InitialiseConservationArrays(self): """ Initialise conservation related vectors/array was in InitialiseModel but has been moved out so is can be called by when the stoichiometry is reanalysed """ # Init and Sx vectors # these vectors are multiplying all by themselves - brett self.__SI__ = numpy.zeros((self.__lzeromatrix__.shape[1]),'d') self.__SALL__ = numpy.zeros((len(self.__species__)),'d') self.__settings__["pitcon_max_step"] = 10*(len(self.__SI__)+1) #max corrector steps if self.__HAS_MOIETY_CONSERVATION__ == True: # reorder the conservation matrix so that it is compatible with L idx = [list(self.conservation_matrix_col).index(n) for n in range(len(self.conservation_matrix_col))] self.__reordered_lcons = self.conservation_matrix.take(idx,1) if self.__HAS_COMPARTMENTS__ and self.__KeyWords__['Species_In_Conc']: self.__Build_Tvec__(amounts=False) else: self.__Build_Tvec__(amounts=True) self.showConserved(screenwrite=0)
[docs] def InitialiseOldFunctions(self): """ InitialiseOldFunctions() Parse and initialise user defined functions specified by !T !U in the PSC input file Arguments: None """ self.__CODE_user = compile(self._Function_user,'_Function_user','exec') try: exec(self.__CODE_user) #print 'done.' except Exception, e: print 'WARNING: User function error\n', e print 'This might be due to non-instantiated (e.g. MCA/state) attributes' print 'Make sure the attributes that are used in your function exist ...' print 'and manually load using the self.ReloadUserFunc() method' #print '\nInitializing init function ...' #print self._Function_init try: self.ReloadInitFunc() except Exception, e: print 'WARNING: Init function error\n', e print 'This function is meant to be used exclusively for the initialization' print 'of PySCeS properties and expressions based on model attributes defined in the input file.' print 'Reinitialize with ReloadInitFunc()'
[docs] def ReloadUserFunc(self): """ ReloadUserFunc() Recompile and execute the user function (!U) from the input file. Arguments: None """ ## print "User functions disabled ..." self.__CODE_user = compile(self._Function_user,'_Function_user','exec') try: exec(self.__CODE_user) except Exception, e: print 'WARNING: User function load error\n', e
[docs] def ReloadInitFunc(self): """ ReloadInitFunc() Recompile and execute the user initialisations (!I) as defined in the PSC input file. and in mod.__InitFuncs__. UPDATE 2015: can now be used to define InitialAssignments (no need for self.* prefix in input file) Arguments: None """ def topolgical_sort(graph_unsorted): """ Repeatedly go through all of the nodes in the graph, moving each of the nodes that has all its edges resolved, onto a sequence that forms our sorted graph. A node has all of its edges resolved and can be moved once all the nodes its edges point to, have been moved from the unsorted graph onto the sorted one. """ # This is the list we'll return, that stores each node/edges pair # in topological order. graph_sorted = [] # Convert the unsorted graph into a hash table. This gives us # constant-time lookup for checking if edges are unresolved, and # for removing nodes from the unsorted graph. graph_unsorted = dict(graph_unsorted) # Run until the unsorted graph is empty. while graph_unsorted: # Go through each of the node/edges pairs in the unsorted # graph. If a set of edges doesn't contain any nodes that # haven't been resolved, that is, that are still in the # unsorted graph, remove the pair from the unsorted graph, # and append it to the sorted graph. Note here that by using # using the items() method for iterating, a copy of the # unsorted graph is used, allowing us to modify the unsorted # graph as we move through it. We also keep a flag for # checking that that graph is acyclic, which is true if any # nodes are resolved during each pass through the graph. If # not, we need to bail out as the graph therefore can't be # sorted. acyclic = False for node, edges in graph_unsorted.items(): for edge in edges: if edge in graph_unsorted: break else: acyclic = True del graph_unsorted[node] graph_sorted.append((node, edges)) if not acyclic: # Uh oh, we've passed through all the unsorted nodes and # weren't able to resolve any of them, which means there # are nodes with cyclic edges that will never be resolved, # so we bail out with an error. raise RuntimeError("A cyclic dependency occurred") return graph_sorted simpleAss = [] exprsAss = [] symbolsX = [] for init in self.__InitFuncs__: if hasattr(self, init): #TODO: to be continued by brett #print(init, self.__InitFuncs__[init]) if type(self.__InitFuncs__[init]) == str: if self.__InitFuncs__[init].isdigit(): #self.__InitFuncs__[init] = eval(self.__InitFuncs__[init]) #setattr(self, init, self.__InitFuncs__[init]) #self.__InitFuncs__[init] = compile(self.__InitFuncs__[init], '_InitFunc_', 'single') simpleAss.append((init, eval(self.__InitFuncs__[init]))) #setattr(self, init, eval(self.__InitFuncs__[init])) else: InfixParser.setNameStr('self.', '') #InfixParser.SymbolReplacements = {'_TIME_':'mod._TIME_'} InfixParser.parse(str(self.__InitFuncs__[init])) piecewises = InfixParser.piecewises symbols = InfixParser.names for s_ in symbols: if s_ not in symbolsX: symbolsX.append(s_) if init not in symbolsX: symbolsX.append(init) functions = InfixParser.functions code_string = 'self.%s = %s' % (init, InfixParser.output) xcode = compile(code_string, '_InitAss_', 'exec') exprsAss.append((init, xcode, symbols)) #setattr(self, init, eval(xcode)) #else: ##setattr(self, init, self.__InitFuncs__[init]) #self.__InitFuncs__[init] = compile(self.__InitFuncs__[init], '_InitFunc_', 'single') #setattr(self, init, eval(self.__InitFuncs__[init])) else: setattr(self, init, self.__InitFuncs__[init]) else: assert hasattr(self, init), '\nModelInit error' #TODO: to be continued by brett #print('symbolsX', symbolsX) for init,val in simpleAss: #print('s', init,val) setattr(self, init, val) execOrder = [] for init, xcode, symbols2 in exprsAss: symbols2 = [symbolsX.index(s_) for s_ in symbols2] execOrder.append((symbolsX.index(init), symbols2)) #TODO: to be continued by brett #print('x', symbols2) #print('e', init,code_string) eval(xcode) #TODO: to be continued by brett #print(execOrder) #print(topolgical_sort(execOrder)) #print(symbolsX)
def __initREQ__(self): """ __initREQ__() Compile and initialise the model rate equations and associated arrays. Rate equations are generated as lambda functions callable as mod.R1(mod) Arguments: None """ # create work arrays for Reactions and variable species # s = numpy.zeros(len(self.__species__),'d') # self.Vtemp = numpy.zeros(len(self.__reactions__),'d') # This appears to be redundant leftover code - brett 20031215 if self.__settings__['display_debug'] == 1: pass #print 'Vtemp' # create the map string by taking the VarReagent[] and assigning it to an s[index] # a reverse map function mapping self.sXi back to self.init[x] for simulation initiation mapString = '' mapString_R = '' for x in range(0,len(self.__species__)): mapString += 'self.%s = s[%s]\n' % (self.__species__[x], x) mapString_R += 'self.__inspec__[%s] = self.%s_init\n' % (x, self.__species__[x]) ## print mapString_R # this creates the mapping string for the derivative functions if self.__settings__['display_debug'] == 1: print 'mapString' print mapString print 'mapString_R' print mapString_R # create the REq string in ReactionIDs order as a single multiline definition # using indexes: vString (Vtemp[x] =) # or a list of individual compile definitions: vArray (v + ReactionID = ) vString = '' ## vString2 = '' ## DvOrder = [] EStat = [] ## dispREq = '' symbR = {} if len(self.__rules__.keys()) > 0 and self.__settings__['mode_substitute_assignment_rules']: # create substitution dictionary for ass in self.__rules__: symbR.update({self.__rules__[ass]['name'] : self.__rules__[ass]['code_string']}) for x in range(0,len(self.__reactions__)): req1 = self.__nDict__[self.__reactions__[x]]['RateEq'] ## DvOrder.append(self.__reactions__[x]) vString += 'Vtemp[%s] = self.%s()\n' % (x, self.__reactions__[x]) # Core update inspired by Core2, lambda functions replaced by Reaction instances if len(self.__rules__.keys()) > 0 and self.__settings__['mode_substitute_assignment_rules']: # substitute assignment rules InfixParser.setNameStr('self.', '') InfixParser.FunctionReplacements = symbR InfixParser.parse(req1.replace('self.', '')) req2 = InfixParser.output ## req1_names = InfixParser.names rObj = ReactionObj(self, self.__reactions__[x], req2, 'self.') else: rObj = ReactionObj(self, self.__reactions__[x], req1, 'self.') self.__ParsePiecewiseFunctions__(rObj.piecewises) rObj.compartment = self.__nDict__[rObj.name]['compartment'] setattr(self, self.__reactions__[x], rObj) # this is to check that if a model defines compartments then REq's MUST be in amounts brett2008 # brett2008 if self.__HAS_COMPARTMENTS__: cnames = [self.__compartments__[c]['name'] for c in self.__compartments__] warnings = '' for rr in self.__reactions__: rrobj = getattr(self, rr) warn = False if rrobj.compartment == None and self.__settings__['display_compartment_warnings']: warnings += "# %s is not located in a compartment.\n" % rrobj.name else: for comp in cnames: if comp in rrobj.symbols: warn = False break else: warn = True if warn: warnings += "# %s: %s\n# assuming kinetic constants are flow constants.\n" % (rr, rrobj.formula) if warnings != '' and self.__settings__['display_compartment_warnings']: print '\n# -- COMPARTMENT WARNINGS --' print warnings os.chdir(self.ModelOutput) if self.__settings__['display_debug'] == 1: print 'vString' print vString ## print 'vString2' ## print vString2 ## print 'DvOrder' ## print DvOrder return mapString,mapString_R,vString def __initmodel__(self): """ __initmodel__() Generate the stoichiometric matrix N from the parsed model description. Returns a stoichiometric matrix (N) Arguments: None """ VarReagents = ['self.'+s for s in self.__species__] StoicMatrix = numpy.zeros((len(VarReagents),len(self.__reactions__)),'d') for reag in VarReagents: for id in self.__reactions__: if reag in self.__nDict__[id]['Reagents'].keys(): StoicMatrix[VarReagents.index(reag)][self.__reactions__.index(id)] = self.__nDict__[id]['Reagents'][reag] return StoicMatrix def __initvar__(self): """ __initvar__() Compile and initialise the model variable species and derivatives and associated mapping arrays. Arguments: None """ mvarString = '' sOrder = [] self.__remaps = '' DvarUpString = '' #self.__inspec__ = numpy.zeros((len(self.__species__)),'d') moved to ParseModel for x in range(0,len(self.__species__)): key = self.__species__[x] ## self.__inspec__[x] = eval(key) self.__inspec__[x] = getattr(self, key) mvarString += 'self.__inspec__[%s] = float(self.%s)\n' % (x, key) self.__remaps += 'self.%s = self.%s_ss\n' % (key, key) sOrder.append('self.'+key) DvarUpString += 'self.%s = input[%s]\n' % (self.__species__[x], x) if self.__settings__['display_debug'] == 1: print 's_initDeriv' print s_initDeriv print '\n__remaps' print self.__remaps print '\nDvarUpString' print DvarUpString mvarFunc = compile(mvarString,'mvarString','exec') return sOrder,DvarUpString def __initfixed__(self): """ __initfixed__() Compile and initialise the fixed species and associated mapping arrays Arguments: None """ runmapString = '' if self.__settings__['display_debug'] == 1: print 'InitParams2' print self.__parameters__ print 'InitStrings2' ## print self.__InitStrings # Initialise parameter elasticities par_map2store = '' par_remap = '' for x in range(len(self.__parameters__)): par_map2store += 'parVal_hold[%s] = self.%s\n' % (x, self.__parameters__[x]) par_remap += 'self.%s = parVal_hold[%s]\n' % (self.__parameters__[x], x) if self.__settings__['display_debug'] == 1: print 'par_map2store' print par_map2store print 'par_remap' print par_remap return (par_map2store,par_remap) #pysces core - the steady-state solver and integration routines def _SpeciesAmountToConc(self, s): """takes and returns s""" for x in range(len(self.__CsizeAllIdx__)): self.__CsizeAll__[x] = getattr(self, self.__CsizeAllIdx__[x]) ## print '__CALL__', self.__CsizeAll__ return s/self.__CsizeAll__ def _SpeciesConcToAmount(self, s): """takes and returns s""" for x in range(len(self.__CsizeAllIdx__)): self.__CsizeAll__[x] = getattr(self, self.__CsizeAllIdx__[x]) ## print '__CALL__', self.__CsizeAll__ return s*self.__CsizeAll__ def _FixedSpeciesAmountToConc(self): for x in range(len(self.__fixed_species__)): am = getattr(self, self.__fixed_species__[x]) self.__CsizeFS__[x] = getattr(self, self.__CsizeFSIdx__[x]) setattr(self, self.__fixed_species__[x], am/self.__CsizeFS__[x]) def _FixedSpeciesConcToAmount(self): for x in range(len(self.__fixed_species__)): cn = getattr(self, self.__fixed_species__[x]) self.__CsizeFS__[x] = getattr(self, self.__CsizeFSIdx__[x]) setattr(self, self.__fixed_species__[x], cn*self.__CsizeFS__[x]) # extract SI and "correct" SD in s solve for v # Vtemp is passed through the function to avoid an irritating bug # that appears if it isn't (ie defined as a global) - brett def _EvalREq2_alt(self,s,Vtemp): """ _EvalREq2_alt(s,Vtemp) Evaluate the rate equations correcting for mass conservation. Takes full [s] returns full [s] --> moiety aware Arguments: s: a vector of species cosnservations Vtemp: the rate vector """ #form an SI vector for x in range(0,len(self.lzeromatrix_col)): self.__SI__[x] = s[self.lzeromatrix_col[x]] #replace SD with SI calculated values for x in range(0,len(self.lzeromatrix_row)): s[self.lzeromatrix_row[x]] = self.__tvec_a__[x] + numpy.add.reduce(self.__lzeromatrix__[x,:]*self.__SI__) return self._EvalREq(s,Vtemp) def _EvalREq(self,s,Vtemp): if self.__HAS_RATE_RULES__: exec(self.__CODE_raterule_map) if self.__HAS_COMPARTMENTS__ and self.__KeyWords__['Species_In_Conc']: s = self._SpeciesAmountToConc(s) exec(self.__mapFunc__) try: self.Forcing_Function() except Exception, de: print 'INFO: forcing function failure', de try: exec(self.__vFunc__) except (ArithmeticError,AttributeError,NameError,ZeroDivisionError,ValueError), detail: print 'INFO: REq evaluation failure:', detail Vtemp[:] = self.__settings__['mach_floateps'] if self.__HAS_RATE_RULES__: try: exec(self.__CODE_raterule) except (ArithmeticError,AttributeError,NameError,ZeroDivisionError,ValueError), detail: print 'INFO: RateRule evaluation failure:', detail return Vtemp def _EvalREq2(self,s,Vtemp): """ _EvalREq2(s,Vtemp) Evaluate the rate equations, as PySCeS uses the reduced set of ODE's for its core operations, this method takes mass conservation into account and regenerates the full species vector takes [si] returns full [s] Arguments: s: species vector Vtemp: rate vector """ #stick SI into s using Nrrow order for x in range(len(s)): self.__SALL__[self.nrmatrix_row[x]] = s[x] #stick SD into s using lzerorow order for x in range(len(self.lzeromatrix_row)): self.__SALL__[self.lzeromatrix_row[x]] = self.__tvec_a__[x] + numpy.add.reduce(self.__lzeromatrix__[x,:]*s) return self._EvalREq(self.__SALL__,Vtemp) def _EvalODE(self,s,Vtemp): """ _EvalODE(s,Vtemp) Core ODE evaluation routine evaluating the reduced set of ODE's. Depending on whether mass conservation is present or not either N*v (_EvalREq) or Nr*v (_EvalREq2) is used to automatically generate the ODE's. Returns sdot. Arguments: s: species vector Vtemp: rate vector """ if self.__HAS_RATE_RULES__: s, self.__rrule__ = numpy.split(s,[self.Nrmatrix.shape[0]]) if self.__HAS_MOIETY_CONSERVATION__ == True: for x in range(len(s)): self.__SALL__[self.nrmatrix_row[x]] = s[x] for x in range(len(self.lzeromatrix_row)): self.__SALL__[self.lzeromatrix_row[x]] = self.__tvec_a__[x] + numpy.add.reduce(self.__lzeromatrix__[x,:]*s) self.__vvec__ = self._EvalREq(self.__SALL__,Vtemp) s = numpy.add.reduce(self.__nrmatrix__*self.__vvec__,1) else: self.__vvec__ = self._EvalREq(s,Vtemp) s = numpy.add.reduce(self.__nmatrix__*self.__vvec__,1) if self.__HAS_RATE_RULES__: s = numpy.concatenate([s, self.__rrule__]).copy() return s # the following are user functions defined in the input file or assigned after instantiated def _EvalODE_CVODE(self,s,Vtemp): """ _EvalODE_CVODE(s,Vtemp) Evaluate the full set of ODE's. Depending on whether mass conservation is present or not either N*v (_EvalREq) or Nr*v (_EvalREq2_alt) is used. Arguments: s: species vector Vtemp: rate vector """ if self.__HAS_RATE_RULES__: s, self.__rrule__ = numpy.split(s,[self.Nmatrix.shape[0]]) self.__vvec__ = self._EvalREq(s,Vtemp) s = numpy.add.reduce(self.__nmatrix__*self.__vvec__,1) if self.__HAS_RATE_RULES__: s = numpy.concatenate([s, self.__rrule__]).copy() return s
[docs] def Forcing_Function(self): """ Forcing_Function() User defined forcing function either defined in the PSC input file as !F or by overwriting this method. This method is evaluated prior to every rate equation evaluation. Arguments: None """ # initialized in __initFunction__() - brett 20050621 exec(self.__CODE_forced)
[docs] def User_Function(self): """ **Deprecated** """ exec(self.__CODE_user) # pysundials
def _EvalExtraData(self, xdata): """ Takes a list of model attributes and returns an array of values """ return numpy.array([getattr(self, d) for d in xdata]) __CVODE_initialise__ = True CVODE_continuous_result = None __CVODE_initial_num__ = None
[docs] def CVODE_continue(self, tvec): """ **Experimental:** continues a simulation over a new time vector, the CVODE memobj is reused and not reinitialised and model parameters can be changed between calls to this method. The mod.data_sim objects from the initial simulation and all calls to this method are stored in the list *mod.CVODE_continuous_result*. - *tvec* a numpy array of time points """ assert self.mode_integrator == 'CVODE', "\nFor what should be rather obvious reasons, this method requires CVODE to be used as the default integration algorithm.\n" if self.CVODE_continuous_result == None: self.CVODE_continuous_result = [self.data_sim] self.__CVODE_initialise__ = False self.sim_time = tvec Tsim0 = time.time() sim_res, rates, simOK = self.CVODE(None) Tsim1 = time.time() print "%s time for %s points: %s" % (self.mode_integrator, len(self.sim_time), Tsim1-Tsim0) if self.__HAS_RATE_RULES__: sim_res, rrules = numpy.split(sim_res,[len(self.__species__)],axis=1) print 'RateRules evaluated and added to mod.data_sim.' # TODO: split this off into a method shared by this and Simulate() self.data_sim = IntegrationDataObj() self.IS_VALID = simOK self.data_sim.setTime(self.sim_time) self.data_sim.setSpecies(sim_res, self.__species__) self.data_sim.setRates(rates, self.__reactions__) if self.__HAS_RATE_RULES__: self.data_sim.setRules(rrules, self.__rate_rules__) if len(self.CVODE_extra_output) > 0: self.data_sim.setXData(self.CVODE_xdata, lbls=self.CVODE_extra_output) self.CVODE_xdata = None if not simOK: print 'Simulation failure' del sim_res self.CVODE_continuous_result.append(self.data_sim) self.__CVODE_initialise__ = True
[docs] def CVODE(self, initial): """ CVODE(initial) PySCeS interface to the CVODE integration algorithm. Arguments: initial: vector containing initial species concentrations """ assert _HAVE_PYSUNDIALS, '\nPySundials is not installed or did not import correctly\n%s' % _PYSUNDIALS_LOAD_ERROR Vtemp = numpy.zeros((self.__Nshape__[1])) def findi(t, y, ydot, f_data): self._TIME_ = t ydota = self._EvalODE(numpy.array(y), Vtemp) ydot[:] = ydota[:] return 0 #non-zero return indicates error state def ffull(t, y, ydot, f_data): self._TIME_ = t ## ya = numpy.array(y) ydota = self._EvalODE_CVODE(numpy.array(y), Vtemp) # unreduced ODE's ydot[:] = ydota[:] return 0 func = None if self.mode_integrate_all_odes: func = ffull else: func = findi if self.__CVODE_initialise__: tZero = initial.copy() if self.__HAS_RATE_RULES__: initial, rrules = numpy.split(initial,[self.Nrmatrix.shape[0]]) tZero = initial.copy() if self.__HAS_MOIETY_CONSERVATION__ and self.mode_integrate_all_odes: initial = self.Fix_S_indinput(initial, amounts=True) tZero = initial.copy() elif self.__HAS_MOIETY_CONSERVATION__: tZero = self.Fix_S_indinput(tZero, amounts=True) if self.__HAS_RATE_RULES__: initial = numpy.concatenate([initial, rrules]) tZero = numpy.concatenate([tZero, rrules]) else: tZero = None #the following block initialises the cvode integrator, and sets various options if self.__CVODE_initialise__: self.__CVODE_y__ = cvode.NVector(initial.tolist()) self.__CVODE_mem__ = cvode.CVodeCreate(cvode.CV_BDF, cvode.CV_NEWTON) #initialisation with basic options, newtonian solver etc... self.__CVODE_initial_num__ = len(initial) del initial t = cvode.realtype(0.0) rates = numpy.zeros((len(self.sim_time), len(self.__reactions__))) output = None if self.__HAS_RATE_RULES__: output = numpy.zeros((len(self.sim_time), len(self.__rrule__)+len(self.__species__))) else: output = numpy.zeros((len(self.sim_time), len(self.__species__))) CVODE_XOUT = False if len(self.CVODE_extra_output) > 0: out = [] for d in self.CVODE_extra_output: if hasattr(self, d) and d not in self.__species__ + self.__reactions__ + self.__rate_rules__: out.append(d) else: print '\nWarning: CVODE is ignoring extra data (%s), it either doesn\'t exist or it\'s a species or rate.\n' % d if len(out) > 0: self.CVODE_extra_output = out CVODE_XOUT = True del out if CVODE_XOUT: self.CVODE_xdata = numpy.zeros((len(self.sim_time), len(self.CVODE_extra_output))) sim_st = 0 if self.sim_time[0] == 0.0: if self.__CVODE_initialise__: output[0,:] = tZero[:].copy() else: output[0,:] = numpy.array(self.__CVODE_y__[:]) if CVODE_XOUT: self.CVODE_xdata[0,:] = self._EvalExtraData(self.CVODE_extra_output) sim_st = 1 del tZero var_store = {} if self.__HAS_EVENTS__: for ev in self.__events__: ev.reset() for ass in ev.assignments: var_store.update({ass.variable : getattr(self, ass.variable)}) TOL_ADJUSTER = 0 MAX_TOL_CNT = 5 MAX_REL_TOLERANCE = 1.0e-3 RELTOL_ADJUST_FACTOR = 1.0e3 MIN_ABS_TOL = self.__settings__["cvode_abstol"] # 1.0e-15 ## MAX_ABS_TOL = self.__settings__["cvode_abstol_max"] #1.0e-3 not used anymore ABSTOL_ADJUST_FACTOR = self.__settings__["cvode_abstol_factor"] # 1.0e-6 cvode_sim_range = range(sim_st, len(self.sim_time)) cvode_scale_range = range(sim_st, len(self.sim_time), len(self.sim_time)/4 or len(self.sim_time)) cvode_scale_range = cvode_scale_range[1:] reltol = cvode.realtype(self.__settings__["cvode_reltol"]) #relative tolerance must be a realtype abstol = cvode.NVector(self.__CVODE_initial_num__*[self.__settings__["cvode_abstol"]]) for s in range(len(self.__CVODE_y__)): newVal = abs(self.__CVODE_y__[s])*ABSTOL_ADJUST_FACTOR if newVal < MIN_ABS_TOL: abstol[s] = MIN_ABS_TOL else: abstol[s] = newVal if self.__CVODE_initialise__: cvode.CVodeMalloc(self.__CVODE_mem__, func, 0.0, self.__CVODE_y__, cvode.CV_SV, reltol, abstol) #set tolerances, specify vector of initial conditions, set integrtion function etc... cvode.CVDense(self.__CVODE_mem__, self.__CVODE_initial_num__) #set dense option, specify dimension of problem (3) if self.__HAS_EVENTS__: cvode.CVodeRootInit(self.__CVODE_mem__, len(self.__events__), self.CVODE_EVENTS, None) #specify to 'root' conditions, and function that calculates them for st in cvode_sim_range: tout = self.sim_time[st] errcount = 0 ## print TOL_ADJUSTER, MAX_TOL_CNT, abstol, MAX_REL_TOLERANCE if self.__settings__["cvode_auto_tol_adjust"] and TOL_ADJUSTER >= MAX_TOL_CNT and reltol.value < MAX_REL_TOLERANCE: reltol.value = reltol.value*RELTOL_ADJUST_FACTOR cvode.CVodeSetTolerances(self.__CVODE_mem__, cvode.CV_SV, reltol, abstol) ## print '\nCVODE: new tolerance set:\nreltol=%s' % reltol.value ## print '\nAbs tolerance:\n%s' % abstol TOL_ADJUSTER = 0 if (st in cvode_scale_range) and self.__settings__["cvode_auto_tol_adjust"]: for s in range(len(self.__CVODE_y__)): newVal = abs(self.__CVODE_y__[s])*ABSTOL_ADJUST_FACTOR if newVal < MIN_ABS_TOL: abstol[s] = MIN_ABS_TOL else: abstol[s] = newVal ## print '\nCVODE: new tolerance set, abstol:\n%s' % abstol cvode.CVodeSetTolerances(self.__CVODE_mem__, cvode.CV_SV, reltol, abstol) ## cvode.CVodeReInit(self.__CVODE_mem__, func, self.sim_time[st-1], self.__CVODE_y__, cvode.CV_SV, reltol, abstol) while True: try: flag = cvode.CVode(self.__CVODE_mem__, tout, self.__CVODE_y__, cvode.ctypes.byref(t), cvode.CV_NORMAL) except AssertionError, ex: print 'cvode error1', ex flag = None self._TIME_ = tout if flag == cvode.CV_ROOT_RETURN: #if a root was found before desired time point, output it ya = numpy.array(self.__CVODE_y__) rootsfound = cvode.CVodeGetRootInfo(self.__CVODE_mem__, len(self.__events__)) reInit = False for ev in range(len(self.__events__)): if rootsfound[ev] == 1: for ass in self.__events__[ev].assignments: # only can assign to independent species vector if ass.variable in self.L0matrix.getLabels()[1] or\ (self.mode_integrate_all_odes and ass.variable in self.__species__): assVal = ass.getValue() assIdx = self.__species__.index(ass.variable) if self.__KeyWords__['Species_In_Conc']: ## print self.__CVODE_y__ self.__CVODE_y__[assIdx] = assVal*getattr(self, self.__CsizeAllIdx__[assIdx]) ## raw_input(self.__CVODE_y__) else: self.__CVODE_y__[assIdx] = assVal reInit = True elif not self.mode_integrate_all_odes and ass.variable in self.L0matrix.getLabels()[0]: print 'Event assignment to dependent species consider setting \"mod.mode_integrate_all_odes = True\"' elif self.__HAS_RATE_RULES__ and ass.variable in self.__rate_rules__: ## print 'Event is assigning to rate rule' assVal = ass.getValue() rrIdx = self.__rate_rules__.index(ass.variable) ## print ass.variable, assVal self.__rrule__[rrIdx] = assVal ## print self.L0matrix.shape[1], rrIdx, len(self.__CVODE_y__) self.__CVODE_y__[self.L0matrix.shape[1]+rrIdx] = assVal setattr(self, ass.variable, assVal) reInit = True else: try: setattr(self, ass.variable, ass.getValue()) reInit = True except: print 'ERROR: Updating model attribute from event: ', ass.variable if reInit: cvode.CVodeReInit(self.__CVODE_mem__, func, tout, self.__CVODE_y__, cvode.CV_SV, reltol, abstol) # this gets everything into the current tout state tmp = None if not self.mode_integrate_all_odes: tmp = self._EvalODE(ya.copy(),self.__CVODE_Vtemp) else: tmp = self._EvalODE_CVODE(ya.copy(),self.__CVODE_Vtemp) del tmp # here we regenerate Sd's and fix concentrations rrules = None if self.__HAS_MOIETY_CONSERVATION__ and not self.mode_integrate_all_odes: if self.__HAS_RATE_RULES__: ya, rrules = numpy.split(ya,[self.Nrmatrix.shape[0]]) ya = self.Fix_S_indinput(ya, amounts=True) # convert to concentrations if self.__HAS_COMPARTMENTS__ and self.__KeyWords__['Output_In_Conc']: ya = self._SpeciesAmountToConc(ya) if self.__HAS_RATE_RULES__: ya = numpy.concatenate([ya, rrules]) else: if self.__HAS_RATE_RULES__: ya, rrules = numpy.split(ya,[self.Nmatrix.shape[0]]) if self.__HAS_COMPARTMENTS__ and self.__KeyWords__['Output_In_Conc']: ya = self._SpeciesAmountToConc(ya) if self.__HAS_RATE_RULES__: ya = numpy.concatenate([ya, rrules]) output[st] = ya # set with self._EvalODE above rates[st] = self.__vvec__ if CVODE_XOUT: self.CVODE_xdata[st,:] = self._EvalExtraData(self.CVODE_extra_output) # this should adjust the expected time to the new output time time self.sim_time[st] = float(tout) # dont need anymore i think if _HAVE_VPYTHON: self.CVODE_VPYTHON(ya) del ya, rrules break if flag == cvode.CV_SUCCESS: ya = numpy.array(self.__CVODE_y__) # this gets everything into the current tout state tmp = None if not self.mode_integrate_all_odes: tmp = self._EvalODE(ya.copy(),self.__CVODE_Vtemp) else: tmp = self._EvalODE_CVODE(ya.copy(),self.__CVODE_Vtemp) del tmp # here we regenerate Sd's and fix concentrations rrules = None if self.__HAS_MOIETY_CONSERVATION__ and not self.mode_integrate_all_odes: if self.__HAS_RATE_RULES__: ya, rrules = numpy.split(ya,[self.Nrmatrix.shape[0]]) ya = self.Fix_S_indinput(ya, amounts=True) # convert to concentrations if self.__HAS_COMPARTMENTS__ and self.__KeyWords__['Output_In_Conc']: ya = self._SpeciesAmountToConc(ya) if self.__HAS_RATE_RULES__: ya = numpy.concatenate([ya, rrules]) else: if self.__HAS_RATE_RULES__: ya, rrules = numpy.split(ya,[self.Nmatrix.shape[0]]) if self.__HAS_COMPARTMENTS__ and self.__KeyWords__['Output_In_Conc']: ya = self._SpeciesAmountToConc(ya) if self.__HAS_RATE_RULES__: ya = numpy.concatenate([ya, rrules]) output[st] = ya # set with self._EvalODE above rates[st] = self.__vvec__ if CVODE_XOUT: self.CVODE_xdata[st,:] = self._EvalExtraData(self.CVODE_extra_output) if _HAVE_VPYTHON: self.CVODE_VPYTHON(ya) del ya, rrules break elif flag == -1: if self.__settings__["cvode_mxstep"] == 1000: self.__settings__["cvode_mxstep"] = 3000 TOL_ADJUSTER += 1 elif self.__settings__["cvode_mxstep"] == 3000: self.__settings__["cvode_mxstep"] = 10000 TOL_ADJUSTER += 2 elif self.__settings__["cvode_mxstep"] == 10000: ## TOL_ADJUSTER += 1 output[st] = numpy.NaN break print 'mxstep warning (%s) mxstep set to %s' % (flag, self.__settings__["cvode_mxstep"]) cvode.CVodeSetMaxNumSteps(self.__CVODE_mem__, self.__settings__["cvode_mxstep"]) elif flag < -3: print 'CVODE error:', flag print 'At ', tout output[st] = numpy.NaN rates[st] = numpy.NaN if CVODE_XOUT: self.CVODE_xdata[st] = numpy.NaN break self.__settings__["cvode_mxstep"] = 1000 cvode.CVodeSetMaxNumSteps(self.__CVODE_mem__, self.__settings__["cvode_mxstep"]) if self.__HAS_EVENTS__: for ass in var_store.keys(): #print 'old value', ass, getattr(self, ass) setattr(self, ass, var_store[ass]) #print 'new value', getattr(self, ass) if self.__settings__["cvode_stats"]: #print some stats from the intgrator nst = cvode.CVodeGetNumSteps(self.__CVODE_mem__) nfe = cvode.CVodeGetNumRhsEvals(self.__CVODE_mem__) nsetups = cvode.CVodeGetNumLinSolvSetups(self.__CVODE_mem__) netf = cvode.CVodeGetNumErrTestFails(self.__CVODE_mem__) nni = cvode.CVodeGetNumNonlinSolvIters(self.__CVODE_mem__) ncfn = cvode.CVodeGetNumNonlinSolvConvFails(self.__CVODE_mem__) nje = cvode.CVDenseGetNumJacEvals(self.__CVODE_mem__) nfeLS = cvode.CVDenseGetNumRhsEvals(self.__CVODE_mem__) nge = cvode.CVodeGetNumGEvals(self.__CVODE_mem__) print "\nFinal Statistics:" print "nst = %-6i nfe = %-6i nsetups = %-6i nfeLS = %-6i nje = %i"%(nst, nfe, nsetups, nfeLS, nje) print "nni = %-6ld ncfn = %-6ld netf = %-6ld nge = %ld\n "%(nni, ncfn, netf, nge) print 'reltol = %s' % reltol print 'abstol:\n%s' % abstol if cvode.CV_SUCCESS >= 0: return output, rates, True else: return output, rates, False
[docs] def CVODE_EVENTS(self, t, svec, eout, f_data): svec = numpy.array(svec) self._TIME_ = t rrules = [] tmp = None if not self.mode_integrate_all_odes: tmp = self._EvalODE(svec.copy(),self.__CVODE_Vtemp) else: tmp = self._EvalODE_CVODE(svec.copy(),self.__CVODE_Vtemp) del tmp eventFired = False for ev in range(len(self.__events__)): if self.__events__[ev](t): eout[ev] = 0 eventFired = True else: eout[ev] = 1 if self.__HAS_RATE_RULES__: exec(self.__CODE_raterule) return 0
[docs] def CVODE_VPYTHON(self, s): """Future VPython hook for CVODE""" pass
[docs] def LSODA(self,initial): """ LSODA(initial) PySCeS interface to the LSODA integration algorithm. Given a set of initial conditions LSODA returns an array of species concentrations and a status flag. LSODA controls are accessible as mod.lsoda_<control> Arguments: initial: vector containing initial species concentrations """ Vtemp = numpy.zeros((self.__Nshape__[1]),'d') def function_sim(s,t): self._TIME_ = t return self._EvalODE(s,Vtemp) iter = 0 go = True status = 0 while go: sim_res,infodict = scipy.integrate.odeint(function_sim,initial,self.sim_time,\ atol=self.__settings__['lsoda_atol'],\ rtol=self.__settings__['lsoda_rtol'],\ mxstep=self.__settings__["lsoda_mxstep"],\ h0=self.__settings__["lsoda_h0"],\ hmax=self.__settings__["lsoda_hmax"],\ hmin=self.__settings__["lsoda_hmin"],\ mxordn=self.__settings__["lsoda_mxordn"],\ mxords=self.__settings__["lsoda_mxords"],\ printmessg=self.__settings__["lsoda_mesg"],\ full_output=1) if infodict['message'] == 'Integration successful.': status = 0 else: status = 1 if status > 0 and iter < self.__settings__['mode_sim_max_iter']: if self.__settings__["lsoda_mxstep"] == 0: print '\nIntegration error\n\nSetting self.__settings__["lsoda_mxstep"] = 1000 and reSimulating ...' self.__settings__["lsoda_mxstep"] = 1000 else: print 'Integration error\n\nSetting self.__settings__["lsoda_mxstep"] = ' + `self.__settings__["lsoda_mxstep"]*3` + ' and reSimulating ...' self.__settings__["lsoda_mxstep"] = self.__settings__["lsoda_mxstep"]*3 iter += 1 elif status > 0 and iter == self.__settings__['mode_sim_max_iter']: print '\nThis simulation is going nowhere fast\nConsider trying CVODE (mod.mode_integrator = \'CVODE\')\n' print 'self.__settings__["lsoda_mxstep"] = ' + `self.__settings__["lsoda_mxstep"]` print '__settings__[\'mode_sim_max_iter\'] = ' + `iter` go = False else: go = False self.__settings__["lsoda_mxstep"] = 0 rates = numpy.zeros((sim_res.shape[0], len(self.__reactions__))) if status == 0: tmp = None for r in range(sim_res.shape[0]): tmp = self._EvalODE(sim_res[r].copy(), self.__CVODE_Vtemp) # set with self._EvalODE above rates[r] = self.__vvec__ del tmp # regenerate dependent variables if self.__HAS_RATE_RULES__: sim_res, rrules = numpy.split(sim_res,[self.Nrmatrix.shape[0]],axis=1) if self.__HAS_MOIETY_CONSERVATION__ == True: res = numpy.zeros((sim_res.shape[0], len(self.__species__))) for x in range(0,sim_res.shape[0]): res[x,:] = self.Fix_S_indinput(sim_res[x,:], amounts=True) sim_res = res del res if self.__HAS_COMPARTMENTS__ and self.__KeyWords__['Output_In_Conc']: for x in range(0,sim_res.shape[0]): sim_res[x] = self._SpeciesAmountToConc(sim_res[x]) if self.__HAS_RATE_RULES__: sim_res = numpy.concatenate([sim_res, rrules],axis=1) return sim_res, rates, True else: if self.__HAS_MOIETY_CONSERVATION__ == True and self.__HAS_RATE_RULES__: sim_res = numpy.zeros((sim_res.shape[0],len(self.__species__)+len(self.__rrule__)),'d') elif self.__HAS_MOIETY_CONSERVATION__ == True: sim_res = numpy.zeros((sim_res.shape[0],len(self.__species__)),'d') sim_res[:] = scipy.NaN return sim_res, rates, False
[docs] def HYBRD(self,initial): """ HYBRD(initial) PySCeS interface to the HYBRD solver. Returns a steady-state solution and error flag. Good general purpose solver. Algorithm controls are available as mod.hybrd_<control> Arguments: initial: vector of initial species concentrations """ Vtemp = numpy.zeros((self.__Nshape__[1]),'d') def function_state(s): return self._EvalODE(s,Vtemp) state_out = scipy.optimize.fsolve(function_state,initial,args=(),\ xtol=self.__settings__['hybrd_xtol'],\ maxfev=self.__settings__['hybrd_maxfev'],\ epsfcn=self.__settings__['hybrd_epsfcn'],\ factor=self.__settings__['hybrd_factor'],\ col_deriv=0,\ full_output=1) if state_out[2] == 1: if self.__settings__['hybrd_mesg'] == 1: print '(hybrd)', state_out[3] return state_out[0], True else: if self.__settings__['hybrd_mesg']: print 'INFO: (hybrd) Invalid steady state:' if self.__settings__['hybrd_mesg'] == 1: print '(hybrd)', state_out[3] return state_out[0], False
[docs] def FINTSLV(self,initial): """ FINTSLV(initial) Forward integration steady-state solver. Finds a steady state when the maximum change in species concentration falls within a specified tolerance. Returns the steady-state solution and a error flag. Algorithm controls are available as mod.fintslv_<control> Arguments: initial: vector of initial concentrations """ sim_time = self.__fintslv_range__*self.__settings__['fintslv_rmult'] #print sim_time Vtemp = numpy.zeros((self.__Nshape__[1]),'d') def function_sim(s,t): return self._EvalODE(s,Vtemp) res,infodict = scipy.integrate.odeint(function_sim,initial.copy(),sim_time,\ atol=self.__settings__['lsoda_atol'],\ rtol=self.__settings__['lsoda_rtol'],\ ## mxstep=self.__settings__["lsoda_mxstep"],\ mxstep=10000,\ h0=self.__settings__["lsoda_h0"],\ hmax=self.__settings__["lsoda_hmax"],\ hmin=self.__settings__["lsoda_hmin"],\ mxordn=self.__settings__["lsoda_mxordn"],\ mxords=self.__settings__["lsoda_mxords"],\ printmessg=self.__settings__["lsoda_mesg"],\ full_output=1) if infodict['message'] == 'Integration successful.': status = True else: status = False # run through results if max(abs([x]-[x-1])) < self.__settings__['fintslv_tol'] score +1 # if you get 5 points by seq end ... happiness OK = 0 if status: for x in range(len(res)): if OK >= self.__settings__['fintslv_step']: break if x > 0 and OK < self.__settings__['fintslv_step']: dif = abs(res[x] - res[x-1]) if max(dif) < self.__settings__['fintslv_tol']: OK += 1 else: pass if OK >= 5: return res[-1], True else: return res[-1], False
[docs] def NLEQ2(self,initial): """ NLEQ2(initial) PySCeS interface to the (optional) NLEQ2 algorithm. This is a powerful steady-state solver that can usually find a solution for when HYBRD() fails. Algorithm controls are available as: mod.nleq2_<control> Returns as steady-state solution and error flag. Arguments: initial: vector of initial species concentrations """ Vtemp = numpy.zeros((self.__Nshape__[1]),'d') initial0 = initial.copy() s_scale = initial.copy() N = len(initial) iwk = numpy.zeros((N+52),'i') rwk = numpy.zeros(((N+max(N,10)+15)*N + 61),'d') iopt = numpy.zeros((50),'i') if self.__settings__['nleq2_jacgen'] == 1: print '(nleq2)User supplied Jacobian not supported yet ... setting __settings__[\'nleq2_jacgen\'] = 0' self.__settings__['nleq2_jacgen'] = 0 rtol = mach_spec.eps*10.0*N iopt[2] = self.__settings__['nleq2_jacgen'] #2 iopt[8] = self.__settings__['nleq2_iscaln'] #0 iopt[10] = 0 iopt[11] = 6 iopt[12] = 0 iopt[13] = 6 iopt[14] = 0 iopt[15] = 6 iopt[30] = self.__settings__['nleq2_nonlin'] #4 iopt[31] = self.__settings__['nleq2_qrank1'] #1 iopt[34] = self.__settings__['nleq2_qnscal'] #0 iopt[37] = self.__settings__['nleq2_ibdamp'] #0 iopt[38] = self.__settings__['nleq2_iormon'] #2 iwk[30] = self.nleq2_nitmax def func(s, ifail): s = self._EvalODE(s,Vtemp) if numpy.isnan(s).any(): ifail = -1 elif (numpy.abs(s)>1.0e150).any(): ifail = -1 return s, ifail def jacfunc(s): return s ierr = 0 BRETT_DEBUG_MODE = False ADVANCED_MODE = self.__settings__['nleq2_advanced_mode'] if ADVANCED_MODE: max_iter = self.__settings__['nleq2_iter'] # 3 max_iter_ceiling = self.__settings__['nleq2_iter_max'] # 10 else: max_iter_ceiling = self.__settings__['nleq2_iter'] max_iter = self.__settings__['nleq2_iter'] iter = 1 GO = True while GO: if BRETT_DEBUG_MODE: print 's_scale', s_scale print 'ierr(%s) = %s' % (iter,ierr) print 'rtol(%s) = %s' % (iter,rtol) print 'nitmax(%s) = %s' % (iter,iwk[30]) print 's_scale(%s) = %s' % (iter,s_scale) print 'max_iter(%s) = %s' % (iter,max_iter) res,s_scale,rtol,iopt,ierr = nleq2.nleq2(func,jacfunc,initial,s_scale,rtol,iopt,iwk,rwk) if ierr == 0: # success GO = False elif ierr == 21: # negative rtol GO = False ## rtol = abs(rtol) ## ierr = 0 elif ierr == 2: # nitmax reached iwk[30] = iwk[30]*self.__settings__['nleq2_growth_factor'] # 10 if iter >= max_iter and iter >= max_iter_ceiling: GO = False iter += 1 if BRETT_DEBUG_MODE and ierr > 0: print 'ierr = %s' % (ierr) print res time.sleep(5) if ierr > 0: if self.__settings__['nleq2_mesg']: print '(nleq2) exits with ierr = %s' % ierr else: if self.__settings__['nleq2_mesg']: print '(nleq2) The solution converged.' if ierr > 0: return res, False else: return res, True
[docs] def PITCON(self,scanpar,scanpar3d=None): """ PITCON(scanpar,scanpar3d=None) PySCeS interface to the PITCON continuation algorithm. Single parameter continuation has been implemented as a "scan" with the continuation being initialised in mod.pitcon_par_space. The second argument does not affect the continuation but can be used to insert a third axis parameter into the results. Returns an array containing the results. Algorithm controls are available as mod.pitcon_<control> Arguments: scanpar: the model parameter to scan (x5) scanpar3d [default=None]: additional output parameter for 3D plots """ if self.__HAS_RATE_RULES__: raise NotImplementedError, '\nBifurcation analysis not currently available for models containing RateRules' assert type(scanpar) == str, '\nscanpar must be a <string> representing a model parameter' modpar = list(self.__parameters__) try: a = modpar.index(scanpar) except: raise NameError, `scanpar` + ' is not a parameter of this model' if scanpar3d != None: if type(scanpar3d) == str: scanpar3d = float(scanpar3d) assert type(scanpar3d) == float, 'scanpar3d must be a <float>' par_hold = getattr(self, scanpar) if self.__settings__["pitcon_jac_opt"] < 1: self.__settings__["pitcon_jac_opt"] = 1 print '\nINFO: .__settings__["pitcon_jac_opt"] set to 1 - user defined jacobian function not yet supported' # DONE! def fx(s): setattr(self, scanpar, s[-1]) try: sdot[:-1] = self._EvalODE(s[:-1], Vtemp) except Exception, ex: print 'PITCON EXCEPTION 1', ex sdot[:-1] = 0.0 sdot[-1] = s[-1] return sdot parm = len(self.__SI__)+1 Vtemp = numpy.zeros((self.__Nshape__[1]),'d') xr2 = numpy.zeros((parm),'d') sdot = numpy.zeros((parm),'d') iwork = numpy.zeros((30 + parm),'i') rwork = numpy.zeros((30 + (6*(parm))*(parm)),'d') ipar = numpy.zeros((parm),'i') fpar = numpy.zeros((parm),'d') res = [] for xscan in self.pitcon_par_space: Vtemp[:] = 0.0 xr2[:] = 0.0 sdot[:] = 0.0 ipar[:] = 0 fpar[:] = 0.0 iwork[0] = 0 #This is a startup iwork[1] = self.__settings__["pitcon_init_par"] #Use X(1) for initial parameter iwork[2] = self.__settings__["pitcon_par_opt"] #Parameterization option 0:allows program iwork[3] = self.__settings__["pitcon_jac_upd"] #Update jacobian every newton step iwork[4] = self.__settings__["pitcon_targ_val_idx"] #Seek target values for X(n) iwork[5] = self.__settings__["pitcon_limit_point_idx"] #Seek limit points in X(n) iwork[6] = self.__settings__["pitcon_output_lvl"] #Control amount of output. iwork[7] = 6 #Output unit 6=PC iwork[8] = self.__settings__["pitcon_jac_opt"] #Jacobian choice. 0:supply jacobian,1:use forward difference,2:central difference iwork[9] = 0 iwork[10] = 0 iwork[11] = 0 iwork[12] = 30 iwork[13] = len(iwork) iwork[14] = 30 + (4*parm) iwork[15] = len(rwork) iwork[16] = self.__settings__["pitcon_max_step"] #max corrector steps iwork[17] = 0 iwork[18] = 0 iwork[19] = 0 iwork[20] = 0 iwork[21] = 0 iwork[22] = 0 iwork[23] = 0 iwork[24] = 0 iwork[25] = 0 iwork[26] = 0 iwork[27] = 0 rwork[0] = self.__settings__["pitcon_abs_tol"] #Absolute error tolerance rwork[1] = self.__settings__["pitcon_rel_tol"] #Relative error tolerance rwork[2] = self.__settings__["pitcon_min_step"] #Minimum stepsize rwork[3] = self.__settings__["pitcon_max_step"] #Maximum stepsize rwork[4] = self.__settings__["pitcon_start_step"] #Starting stepsize rwork[5] = self.__settings__["pitcon_start_dir"] #Starting direction +1.0/-1.0 rwork[6] = self.pitcon_targ_val #Target value (Seek solution with iwork[4]=) rwork[7] = 0.0 rwork[8] = 0.0 rwork[9] = 0.0 rwork[10] = 0.0 rwork[11] = 0.0 rwork[12] = 0.0 rwork[13] = 0.0 rwork[14] = 0.0 rwork[15] = 0.0 rwork[16] = 0.0 rwork[17] = 0.0 rwork[18] = 0.0 rwork[19] = self.__settings__["pitcon_max_grow"] #maximum growth factor rwork[20] = 0.0 rwork[21] = 0.0 rwork[22] = 0.0 rwork[23] = 0.0 rwork[24] = 0.0 rwork[25] = 0.0 rwork[26] = 0.0 rwork[27] = 0.0 rwork[28] = 0.0 setattr(self, scanpar, xscan) self.State() if self.__HAS_COMPARTMENTS__ and self.__KeyWords__['Output_In_Conc']: ## if self.__KeyWords__['Species_In_Conc']: self.__inspec__ = copy.copy(self._SpeciesConcToAmount(self.state_species)) else: self.__inspec__ = copy.copy(self.state_species) go = False if self.__StateOK__: go = True elif not self.__StateOK__ and self.__settings__["pitcon_allow_badstate"]: go = True else: go = False if go: if self.__HAS_MOIETY_CONSERVATION__ == True: temp = numpy.zeros((len(self.__SI__)),'d') #sI0_sim_init for x in range(0,len(self.lzeromatrix_col)): xr2[x] = self.__inspec__[self.nrmatrix_row[x]] else: xr2[:-1] = copy.copy(self.state_species[:]) xr2[-1] = copy.copy(xscan) for x in range(int(self.pitcon_iter)): ierror,iwork,rwork,xr2 = pitcon.pitcon1(fx,fpar,fx,ipar,iwork,rwork,xr2) if iwork[0] == 2: if self.__settings__["pitcon_flux_gen"]: if min(xr2[:-1]) < 0.0 and self.__settings__["pitcon_filter_neg"]: pass else: xout = xr2.tolist() a = xout.pop(-1) if self.__HAS_MOIETY_CONSERVATION__ == True: xout = self.Fix_S_indinput(xout, amounts=True) else: xout = numpy.array(xout) if self.__HAS_COMPARTMENTS__ and self.__KeyWords__['Output_In_Conc']: xout = self._SpeciesAmountToConc(xout) xout2 = (self.__FluxGen__(xout)).tolist() xout = xout.tolist() xout.insert(0,a) if scanpar3d != None: xout.insert(0,scanpar3d) xout2 = xout+xout2 res.append(xout2) else: if min(xr2[:-1]) < 0.0 and self.__settings__["pitcon_filter_neg"]: pass else: xout = xr2.tolist() a = xout.pop(-1) if self.__HAS_MOIETY_CONSERVATION__ == True: xout = self.Fix_S_indinput(xout, amounts=True) else: xout = numpy.array(xout) if self.__HAS_COMPARTMENTS__ and self.__KeyWords__['Output_In_Conc']: xout = self._SpeciesAmountToConc(xout) xout = xout.tolist() xout.insert(0,a) if scanpar3d != None: xout.insert(0,scanpar3d) res.append(xout) elif iwork[0] == 3: print '\nTarget point:' xout = xr2.tolist() a = xout.pop(-1) if self.__HAS_MOIETY_CONSERVATION__ == True: xout = self.Fix_S_indinput(xout, amounts=True) else: xout = numpy.array(xout) if self.__HAS_COMPARTMENTS__ and self.__KeyWords__['Output_In_Conc']: xout = self._SpeciesAmountToConc(xout) xout = xout.tolist() xout.insert(0,a) print xout self.pitcon_target_points.append(xout) elif iwork[0] == 4: print '\nLimit point' xout = xr2.tolist() a = xout.pop(-1) if self.__HAS_MOIETY_CONSERVATION__ == True: xout = self.Fix_S_indinput(xout, amounts=True) else: xout = numpy.array(xout) if self.__HAS_COMPARTMENTS__ and self.__KeyWords__['Output_In_Conc']: xout = self._SpeciesAmountToConc(xout) xout = xout.tolist() xout.insert(0,a) print xout self.pitcon_limit_points.append(xout) elif iwork[0] == 1: pass else: print iwork[0] #raw_input() else: print '\nInvalid steady state, skipping ...' if self.__settings__["pitcon_filter_neg_res"]: for result in range(len(res)-1,-1,-1): if min(res[result][:len(self.species)+1]) < 0.0: res.pop(result) setattr(self, scanpar, par_hold) return numpy.array(res)
[docs] def Simulate(self,userinit=0): """ PySCeS integration driver routine that evolves the system over the time. Resulting array of species concentrations is stored in the **mod.data_sim** object Initial concentrations can be selected using *mod.__settings__['mode_sim_init']* (default=0): - 0 initialise with intial concentrations - 1 initialise with a very small (close to zero) value - 2 initialise with results of previously calculated steady state - 3 initialise with final point of previous simulation *userinit* values can be (default=0): - 0: initial species concentrations intitialised from (mod.S_init), time array calculated from sim_start/sim_end/sim_points - 1: intial species concentrations intitialised from (mod.S_init) existing "mod.sim_time" used directly - 2: initial species concentrations read from "mod.__inspec__", "mod.sim_time" used directly """ # check if the stoichiometry has been adjusted using Stoich_nmatrix_SetValue - brett 20050719 if not self.__StoichOK: self.Stoichiometry_ReAnalyse() # initialises self.__inspec__[x] with self.sXi if userinit == 1: eval(self.__mapFunc_R__) elif userinit == 2: try: assert len(self.__inspec__) == len(self.__species__) except: print '\nINFO: sim_sinit is the incorrect length initialising with .sX_init' self.__inspec__ = numpy.zeros(len(self.__species__)) eval(self.__mapFunc_R__) else: self.sim_start = float(self.sim_start) self.sim_end = float(self.sim_end) self.sim_points = float(self.sim_points) if self.sim_points == 1.0: print '*****\nWARNING: simulations require a minimum of 2 points, setting sim_points = 2.0\n*****' self.sim_points = 2.0 self.sim_time = numpy.linspace(self.sim_start, self.sim_end, self.sim_points, endpoint=1, retstep=0) eval(self.__mapFunc_R__) # initialise __rrule__ to mod.<rule>_init if self.__HAS_RATE_RULES__: for r in range(len(self.__rate_rules__)): self.__rrule__[r] = getattr(self, '%s_init' % self.__rate_rules__[r]) for c in range(len(self.__CsizeAllIdx__)): cval = getattr(self, '%s_init' % self.__CsizeAllIdx__[c]) setattr(self, self.__CsizeAllIdx__[c], cval) self.__CsizeAll__[c] = cval # set initialisation array to amounts if self.__HAS_COMPARTMENTS__ and self.__KeyWords__['Species_In_Conc']: self.__inspec__ = self._SpeciesConcToAmount(self.__inspec__) # set mod.<species> to corrected mod.<species>_init value for s in range(len(self.__species__)): setattr(self, self.__species__[s], self.__inspec__[s]) #This should work ok __Build_Tvec__ uses .self.__inspec__ to create Tvec if self.__HAS_MOIETY_CONSERVATION__ == True: self.__Build_Tvec__(amounts=True) self.showConserved(screenwrite=0) if self.__settings__['display_debug'] == 1: print self.conserved_sums # Initialise the simulation ... if self.__settings__['mode_sim_init'] == 0: # Start with s[x] at their initial values s0_sim_init = copy.copy(self.__inspec__) elif self.__settings__['mode_sim_init'] == 1: # Start with s[x] at zero ... well close to it anyway s0_sim_init = copy.copy(self.__inspec__) s0_sim_init[:] = self.__settings__['small_concentration'] elif self.__settings__['mode_sim_init'] == 2: # Start with s[x] at the previous steady state if it exists and check if s[x] < 0.0 # if so set that value to 1.0e-10 if self.state_species != None: s0_sim_init = copy.copy(self.state_species)*1.0 for x in range(len(s0_sim_init)): if s0_sim_init[x] < 0.0: s0_sim_init[x] = self.__settings__['small_concentration'] print 'Negative concentration detected in SimInit: s['+`x`+'] set to ' + `self.__settings__['small_concentration']` else: s0_sim_init = copy.copy(self.__inspec__) elif self.__settings__['mode_sim_init'] == 3: # Start with s[x] at the final point of the previous simulation if exists -- johann 20050220 try: s0_sim_init = copy.copy(self.data_sim.species[-1]) except: s0_sim_init = copy.copy(self.__inspec__) else: s0_sim_init = copy.copy(self.__inspec__) #print 'Sim debug' #print s0_sim_init if self.__HAS_MOIETY_CONSERVATION__ == True: temp = numpy.zeros((len(self.__SI__)),'d') for x in range(0,len(self.lzeromatrix_col)): temp[x] = s0_sim_init[self.nrmatrix_row[x]] s0_sim_init = temp del temp #print s0_sim_init # ok so now we add RateRules to the initialisation_vec and see what happens if self.__HAS_RATE_RULES__: s0_sim_init = numpy.concatenate([s0_sim_init, self.__rrule__]) # real pluggable integration routines - brett 2007 # the array copy is important ... brett leave it alone! Tsim0 = time.time() if self.mode_integrator == 'LSODA': sim_res, rates, simOK = self.LSODA(copy.copy(s0_sim_init)) elif self.mode_integrator == 'CVODE': sim_res, rates, simOK = self.CVODE(copy.copy(s0_sim_init)) Tsim1 = time.time() print "%s time for %s points: %s" % (self.mode_integrator, len(self.sim_time), Tsim1-Tsim0) if self.__HAS_RATE_RULES__: sim_res, rrules = numpy.split(sim_res,[len(self.__species__)],axis=1) print 'RateRules evaluated and added to mod.data_sim.' self.data_sim = IntegrationDataObj() self.IS_VALID = simOK self.data_sim.setTime(self.sim_time) self.data_sim.setSpecies(sim_res, self.__species__) self.data_sim.setRates(rates, self.__reactions__) if self.__HAS_RATE_RULES__: self.data_sim.setRules(rrules, self.__rate_rules__) if len(self.CVODE_extra_output) > 0: self.data_sim.setXData(self.CVODE_xdata, lbls=self.CVODE_extra_output) self.CVODE_xdata = None if not simOK: print 'Simulation failure' del sim_res
[docs] def State(self): """ State() PySCeS non-linear solver driver routine. Solve for a steady state using HYBRD/NLEQ2/FINTSLV algorithms. Results are stored in mod.state_species and mod.state_flux. The results of a steady-state analysis can be viewed with the mod.showState() method. The solver can be initialised in 3 ways using the mode_state_init switch. mod.mode_state_init = 0 initialize with species initial values mod.mode_state_init = 1 initialize with small values mod.mode_state_init = 2 initialize with the final value of a 10-logstep simulation numpy.logspace(0,5,18) Arguments: None """ # check if the stoichiometry has been adjusted using Stoich_nmatrix_SetValue - brett 20050719 if not self.__StoichOK: self.Stoichiometry_ReAnalyse() self.__StateOK__ = True # function to feed the simulation routine if used to initialise the solver Vtemp = numpy.zeros((self.__Nshape__[1]),'d') def function_sim(s,t): return self._EvalODE(s,Vtemp) #set self.__inspec__ to current Xi values eval(self.__mapFunc_R__) # initialise __rrule__ to mod.<rule>_init if self.__HAS_RATE_RULES__: for r in range(len(self.__rate_rules__)): self.__rrule__[r] = getattr(self, '%s_init' % self.__rate_rules__[r]) # set compartment values to initial values for c in range(len(self.__CsizeAllIdx__)): cval = getattr(self, '%s_init' % self.__CsizeAllIdx__[c]) setattr(self, self.__CsizeAllIdx__[c], cval) self.__CsizeAll__[c] = cval # set initialisation array to amounts if self.__HAS_COMPARTMENTS__ and self.__KeyWords__['Species_In_Conc']: self.__inspec__ = self._SpeciesConcToAmount(self.__inspec__) # set mod.<species> to corrected mod.<species>_init value for s in range(len(self.__species__)): setattr(self, self.__species__[s], self.__inspec__[s]) #This should work ok __Build_Tvec__ uses .self.__inspec__ to create Tvec if self.__HAS_MOIETY_CONSERVATION__ == True: self.__Build_Tvec__(amounts=True) self.showConserved(screenwrite=0) if self.__settings__['display_debug'] == 1: print self.conserved_sums #clear the solver initialisation array s0_ss_init = None # save the __settings__['hybrd_factor'] hybrd_factor_temp = copy.copy(self.__settings__['hybrd_factor']) # use StateInit to initialise the solver with either 0:initval 1:zeroval 2:sim 3:1%state # in all cases fallback to init # initialising to previous ss seems problematic so I use |10%| of previous - brett if self.mode_state_init == 0: # Use initial S values s0_ss_init = self.__inspec__.copy() elif self.mode_state_init == 1: # Use almost zero values 1.0e-8 any smaller seems to cause a problem # this is user defineable option --> model.ZeroVal - brett 20040121 s0_ss_init = self.__inspec__.copy() s0_ss_init[:] = self.__settings__['small_concentration'] elif self.mode_state_init == 2: print 'This initialisation mode has been disabled, using initial values.' s0_ss_init = self.__inspec__.copy() ## # Perform a 10-logstep simulation numpy.logspace(0,5,18) and initialise solver with final value ## # This is guesstimate ... if anyone has a better idea please let me know ## # it should and be a user defineable array (min/max/len)- brett 20031215 ## self.__settings__['hybrd_factor'] = 10 ## try: ## s0_ss_init = scipy.integrate.odeint(function_sim,self.__inspec__.tolist(),self.__mode_state_init2_array__,10000,full_output=0) ## if self.__HAS_MOIETY_CONSERVATION__ == True: ## s0_ss_init = self.Fix_S_fullinput(s0_ss_init[-1], amounts=True) ## else: ## s0_ss_init = s0_ss_init[-1] ## except: ## s0_ss_init = copy.copy(self.__inspec__) elif self.mode_state_init == 3: # Use |10%| of previous steady-state - seems to be safe, needs more testing and probably professional help # My rationale for this is "a set of approximate values where all variables are relatively close" # without having the overhead of a simulation ... # will eventually be a user option so that it can be +/- x% previous steadystate - brett 20031215 # Note: hybrid doesn't seem to like to start too close to its solution more than 10% is dodgy if self.state_species != None and self.__HAS_COMPARTMENTS__ and self.__KeyWords__['Output_In_Conc']: s0_ss_init = self._SpeciesConcToAmount(abs(copy.copy(self.state_species))*float(self.__settings__['mode_state_init3_factor'])) else: s0_ss_init = copy.copy(self.__inspec__) else: s0_ss_init = copy.copy(self.__inspec__) # reset __settings__['hybrd_factor'] self.__settings__['hybrd_factor'] = hybrd_factor_temp #print 'State debug' #print s0_ss_init # form an initialisation vector of independent species if self.__HAS_MOIETY_CONSERVATION__ == True: temp = numpy.zeros((len(self.__SI__)),'d') for x in range(0,len(self.lzeromatrix_col)): temp[x] = s0_ss_init[self.nrmatrix_row[x]] # add RateRules to the initialisation_vec and see what happens if self.__HAS_RATE_RULES__: temp = numpy.concatenate([temp, self.__rrule__]) s0_ss_init = temp del temp #print s0_ss_init available_solvers = ['HYBRD'] # set mode_solver to new syntax for backwards compatibility only if self.mode_solver == 0: self.mode_solver = 'HYBRD' elif self.mode_solver == 1: self.mode_solver = 'FINTSLV' elif self.mode_solver == 2: self.mode_solver = 'NLEQ2' # check for nleq2 and add if available this should be first if nleq2_switch == 1: available_solvers.append('NLEQ2') else: if self.mode_solver == 'NLEQ2': self.mode_solver = 'HYBRD' print 'INFO: switching to HYBRD.\nNleq2 solver not available see /nleq/readme.txt for details' # ******* OTHER SOLVERS GO IN HERE ******* # ******* OTHER SOLVERS GO IN HERE ******* # shall we add HYBRD to fallback? this should be last if self.__settings__['mode_solver_fallback_integration']: available_solvers.append('FINTSLV') if not self.mode_solver_fallback == 1: assert self.mode_solver in available_solvers, '\nERROR: %s is not a valid (%s) solver!' % (solver, str(available_solvers)) available_solvers = [self.mode_solver] # if solver other than HYBRD is selected move it to front of list so it is run first if self.mode_solver != 'HYBRD': available_solvers.insert(0, available_solvers.pop(available_solvers.index(self.mode_solver))) STATE_XOUT = False STATE_xdata = None if len(self.STATE_extra_output) > 0: out = [] for d in self.STATE_extra_output: if hasattr(self, d) and d not in self.__species__ + self.__reactions__ + self.__rate_rules__: out.append(d) else: print '\nWARNING: STATE is ignoring extra data (%s), it either doesn\'t exist or it\'s a species, rate or rule.\n' % d if len(out) > 0: self.STATE_extra_output = out STATE_XOUT = True del out self.__StateOK__ = True state_species = None rrules = None state_flux = None for solver in available_solvers: if solver == 'HYBRD': state_species, self.__StateOK__ = self.HYBRD(s0_ss_init.copy()) elif solver == 'NLEQ2': state_species, self.__StateOK__ = self.NLEQ2(s0_ss_init.copy()) elif solver == 'FINTSLV': state_species, self.__StateOK__ = self.FINTSLV(s0_ss_init.copy()) #In case of a number (scalar) output from the solver if numpy.isscalar(state_species): state_species = numpy.array([state_species],'d') # this gets everything into the current tout state tmp = self._EvalODE(state_species.copy(), Vtemp) state_flux = self.__vvec__ if self.__HAS_RATE_RULES__: state_species, rrules = numpy.split(state_species,[self.Nrmatrix.shape[0]]) # test for negative concentrations if (state_species < 0.0).any(): self.__StateOK__ = False if self.__settings__['mode_state_mesg']: print 'WARNING!! Negative concentrations detected.' if self.__StateOK__: break else: if self.mode_solver_fallback and self.__settings__['solver_switch_warning']: slv_idx = available_solvers.index(solver) if slv_idx != len(available_solvers)-1: print 'INFO: STATE is switching to %s solver.' % available_solvers[slv_idx+1] else: print 'INFO: STATE calculation failed!' if STATE_XOUT: STATE_xdata = self._EvalExtraData(self.STATE_extra_output) # the current status quo is all state algorithms will use and produce results # in amounts and autoconversion to and from species takes place in the calling # algorithms -- brett2008 # THIS IS OPPOSITE TO THE SIMULATE METHOD brett - again if self.__HAS_MOIETY_CONSERVATION__ == True: state_species = self.Fix_S_indinput(state_species, amounts=True) if self.__HAS_COMPARTMENTS__ and self.__KeyWords__['Output_In_Conc']: self.state_species = self._SpeciesAmountToConc(state_species.copy()) else: self.state_species = state_species self.state_flux = state_flux del state_species, state_flux # final check for a bad state set check if fluxes are == mach_eps # this is almost never going to be true if (self.state_flux == self.__settings__['mach_floateps']).any(): print '\nWARNING: extremely small flux detected! proceed with caution:' print self.state_flux ## self.__StateOK__ = False if not self.__StateOK__: print '\n***\nWARNING: invalid steady state solution (species concentrations and fluxes)\n***\n' if self.__settings__['mode_state_nan_on_fail']: self.state_species[:] = numpy.NaN self.state_flux[:] = numpy.NaN # set the instance steady state flux and species attributes self.SetStateSymb(self.state_flux,self.state_species) # coming soon to a terminal near you self.data_sstate = StateDataObj() self.data_sstate.setSpecies(self.state_species, self.__species__) self.data_sstate.setFluxes(self.state_flux, self.__reactions__) if self.__HAS_RATE_RULES__: self.data_sstate.setRules(rrules, self.__rate_rules__) if STATE_XOUT: self.data_sstate.setXData(STATE_xdata, lbls=self.STATE_extra_output) del STATE_xdata self.data_sstate.IS_VALID = self.__StateOK__ if self.__settings__['display_debug'] == 1: print 'self.state_species' print self.state_species print 'self.state_flux' print self.state_flux #driver routines that support the core routines # core support # takes the flux and species arrays and # assigns them as instances variable self.Jx and self.Xss # I'm not sure if anyone wants to use this in real life so it might be hidden at some point - brett 20040122 # gone - brett 20040506 ... back brett 20040720
[docs] def SetStateSymb(self,flux,metab): """ SetStateSymb(flux,metab) Sets the individual steady-state flux and concentration attributes as mod.J_<reaction> and mod.<species>_ss Arguments: flux: the steady-state flux array metab: the steady-state concentration array """ for x in range(0,len(self.state_species)): setattr(self, self.__species__[x]+'_ss', metab[x]) for x in range(0,len(self.state_flux)): setattr(self, 'J_' + self.__reactions__[x], flux[x]) # uses __inspec__ to build Tvec
def __Build_Tvec__(self, amounts=True): """ __Build_Tvec_(concs=False) Creates vectors of conserved moiety totals from __inspec__ self.__tvec_a__ = amounts self.__tvec_c__ = concentrations Arguments: None """ if self.__HAS_MOIETY_CONSERVATION__ == True: if amounts: self.__tvec_a__ = numpy.add.reduce(copy.copy(self.__reordered_lcons)*self.__inspec__,1) self.__tvec_c__ = numpy.add.reduce(copy.copy(self.__reordered_lcons)*self._SpeciesAmountToConc(self.__inspec__),1) else: self.__tvec_c__ = numpy.add.reduce(copy.copy(self.__reordered_lcons)*self.__inspec__,1) self.__tvec_a__ = numpy.add.reduce(copy.copy(self.__reordered_lcons)*self._SpeciesConcToAmount(self.__inspec__),1)
[docs] def showConserved(self,File=None,screenwrite=1,fmt='%2.3f'): """ showConserved(File=None,screenwrite=1,fmt='%2.3f') Print the moiety conserved cycles present in the system. Arguments: File [default=None]: an open writable Python file object screenwrite [default=1]: write results to console (0 means no reponse) fmt [default='%2.3f']: the output number format string """ if self.__HAS_MOIETY_CONSERVATION__ == True: Tlist = range(0,len(self.__tvec_a__)) if Tlist != []: ConSumPstr = '' for x in range(0,len(Tlist)): for y in range (0,len(self.conservation_matrix_col)): if self.conservation_matrix[Tlist[x],y] > 0.0: #coeff = self.__settings__['mode_number_format'] % (s_init[y]*abs(conservation_matrix[Tlist[x],y])) coeff = fmt % abs(self.conservation_matrix[Tlist[x],y]) met = self.__D_s_Order[self.conservation_matrix_col[y]] ConSumPstr += ' + {' + coeff + '}' + met.replace('self.','') elif self.conservation_matrix[Tlist[x],y] < 0.0: #coeff = self.__settings__['mode_number_format'] % (s_init[y]*abs(conservation_matrix[Tlist[x],y])) coeff = fmt % abs(self.conservation_matrix[Tlist[x],y]) met = self.__D_s_Order[self.conservation_matrix_col[y]] ConSumPstr += ' - {' + coeff + '}' + met.replace('self.','') if self.__HAS_COMPARTMENTS__ and self.__KeyWords__['Output_In_Conc']: ConSumPstr += ' = ' + self.__settings__['mode_number_format'] % self.__tvec_c__[Tlist[x]] + '\n' else: ConSumPstr += ' = ' + self.__settings__['mode_number_format'] % self.__tvec_a__[Tlist[x]] + '\n' self.conserved_sums = ConSumPstr else: self.conserved_sums = 'No moiety conservation' if File != None: print '\nConserved relationships' #assert type(File) == file, 'showConserved() needs an open file object' File.write('\n## Conserved relationships\n') File.write(self.conserved_sums) elif screenwrite: print '\nConserved relationships' print self.conserved_sums
[docs] def showFluxRelationships(self,File=None): """ showConserved(File=None) Print the flux relationships present in the system. Arguments: File [default=None]: an open writable Python file object """ Ostr = '' for row in range(self.__kzeromatrix__.shape[0]): Ostr += "%s =" % self.reactions[self.kzeromatrix_row[row]] for col in range(self.__kzeromatrix__.shape[1]): if self.__kzeromatrix__[row,col] != 0.0: if self.__kzeromatrix__[row,col] > 0.0: Ostr += " + {%2.2f}%s" % (abs(self.__kzeromatrix__[row,col]), self.reactions[self.kzeromatrix_col[col]]) else: Ostr += " - {%2.2f}%s" % (abs(self.__kzeromatrix__[row,col]), self.reactions[self.kzeromatrix_col[col]]) Ostr += '\n' if File != None: print '\nFlux relationships' #assert type(File) == file, 'showConserved() needs an open file object' File.write('\n## Flux relationships\n') File.write(Ostr) else: print '\nFlux relationships' print Ostr # Calculate dependant variables done directly in EvalREq2 this is a utility version
[docs] def Fix_S_fullinput(self, s_vec, amounts=True): """ Fix_S_fullinput(s_vec) Using the full concentration vector evaluate the dependent species Arguments: s_vec: a full length concentration vector """ # s_vec = copy.copy(s) for x in range(0,len(self.lzeromatrix_col)): self.__SI__[x] = s_vec[self.lzeromatrix_col[x]] for x in range(0,len(self.lzeromatrix_row)): if amounts: s_vec[self.lzeromatrix_row[x]] = self.__tvec_a__[x] + numpy.add.reduce(self.__lzeromatrix__[x,:]*self.__SI__) #there might be a way to reduce the 2 for loops else: s_vec[self.lzeromatrix_row[x]] = self.__tvec_c__[x] + numpy.add.reduce(self.__lzeromatrix__[x,:]*self.__SI__) #there might be a way to reduce the 2 for loops return s_vec # Calculate dependant variables done directly in EvalREq2B this is a utility version
[docs] def Fix_S_indinput(self, s_vec, amounts=True): """ Fix_S_indinput(s_vec, amounts=True) whether to use self.__tvec_a__ (default) or self.__tvec_c__ Given a vector of independent species evaluate and return a full concentration vector. Arguments: s_vec: vector of independent species """ #stick SI into s using Nrrow order for x in range(len(s_vec)): self.__SALL__[self.nrmatrix_row[x]] = s_vec[x] #stick SD into s using lzerorow order for x in range(len(self.lzeromatrix_row)): if amounts: self.__SALL__[self.lzeromatrix_row[x]] = self.__tvec_a__[x] + numpy.add.reduce(self.__lzeromatrix__[x,:]*s_vec) else: self.__SALL__[self.lzeromatrix_row[x]] = self.__tvec_c__[x] + numpy.add.reduce(self.__lzeromatrix__[x,:]*s_vec) return copy.copy(self.__SALL__) # output support routines # Quick and dirty flux regeneration uses the Rx form of the RE's to cater for potential conservation # caters for both vectors and arrays and is conservation aware
def __FluxGen__(self,s): """ **Deprecating** only used by **PITCON** s: species vector/array """ Vtemp = numpy.zeros((self.__Nshape__[1]),'d') try: s.shape[0] Vout = numpy.zeros((len(s),len(self.__vvec__)),'d') for x in range(len(s)): if self.__HAS_MOIETY_CONSERVATION__ == True: #depending if there is conservation -- N.L_switch is: 0 for none or 1 for present Vout[x,:] = self._EvalREq2_alt(s[x,:],Vtemp) elif self.__HAS_MOIETY_CONSERVATION__ == False: Vout[x,:] = self._EvalREq(s[x,:],Vtemp) except: if self.__HAS_MOIETY_CONSERVATION__ == True: #depending if there is conservation -- N.L_switch is: 0 for none or 1 for present Vout = self._EvalREq2_alt(s,Vtemp) elif self.__HAS_MOIETY_CONSERVATION__ == False: Vout = self._EvalREq(s,Vtemp) return Vout
[docs] def FluxGenSim(self,s): """ **Deprecated** """ pass
[docs] def ParGenSim(self): """ **Deprecated** """ pass
[docs] def Fix_Sim(self,metab,flux=0,par=0): """ **Deprecated** """ pass # MCA routines # elasticity routines
[docs] def EvalEvar(self,input=None,input2=None): """ EvalEvar(input=None,input2=None) Calculate reaction elasticities towards the variable species. Both inputs (input1=species,input2=rates) should be valid (steady state for MCA) solutions and given in the correct order for them to be used. If either or both are missing the last state values are used automatically. Elasticities are scaled using input 1 and 2. Arguments:: - input [default=None]: species concentration vector - input2 [default=None]: reaction rate vector Settings, set in mod.__settings__:: - elas_evar_upsymb [default = 1] attach individual elasticity symbols to model instance - elas_zero_conc_fix [default=False] if zero concentrations are detected in a steady-state solution make it a very small number - elas_zero_flux_fix [default=False] if zero fluxes are detected in a steady-state solution make it a very small number - elas_scaling_div0_fix [default=False] if INf's are detected after scaling set to zero """ if input == None or input2 == None: input = self.state_species input2 = self.state_flux #print 'INFO: Using state_species and state_flux as input' else: assert len(input) == len(self.species), 'length error this array must have ' + str(len(self.species)) + ' elements' assert len(input2) == len(self.reactions), 'length error this array must have ' + str(len(self.reactions)) + ' elements' # not really necessary but in here just in case - brett 20040930 # map input back to self.Sx exec(self.__CDvarUpString__) if self.__settings__['elas_zero_flux_fix']: val = self.__settings__['mach_floateps']**2 for x in range(len(input2)): if abs(input2[x]) <= val: print 'Info: zero flux detected: J_%s set to %s' % (self.reactions[x], val) input2[x] = val if self.__settings__['elas_zero_conc_fix']: val = self.__settings__['mach_floateps']**2 for x in range(len(input)): if abs(input[x]) <= val: print 'Info: zero concentration detected: %s set to %s' % (self.species[x], val) input[x] = val if self.__settings__['display_debug'] == 1: print '\nVarinput = ' + `input` + '\n' self.__evmatrix__ = numpy.zeros((len(self.__reactions__),len(self.__species__)),'d') # scale KL (future refactoring allow user input) self.ScaleKL(input,input2) # create tuples of the rows and columns for the Ev matrix self.elas_var_row = tuple(copy.copy(self.__reactions__)) col = copy.copy(self.__D_s_Order) for x in range(len(self.__D_s_Order)): col[x] = col[x].replace('self.','') self.elas_var_col = tuple(col) del col # attempt to evaluate every variable against every flux: E's > mach_eps are assumed to exist for react in range(len(self.__reactions__)): if self.__settings__['display_debug'] == 1: print '\nReaction: ' + self.__reactions__[react] for met in range(len(self.__species__)): countV = 0 countMet = 0 # this modification should make this independant of a steady state - finally implimented july2004 # eval('self.'+ self.__species__[met] + '_ss'),order=self.__settings__['mode_elas_deriv_order'],dx=hstep,n=1,\ try: """ I got the idea of scaling the stepsize to So from Herbert Sauro's Jarnac TModel.uEEOp function brett - 20040818""" hstep = input[met]*self.__settings__['mode_elas_deriv_factor'] # self.__settings__['mode_elas_deriv_factor'] = 0.0001 if abs(hstep) < self.__settings__['mode_elas_deriv_min']: # self.__settings__['mode_elas_deriv_min'] = 1.0e-12 hstep = self.__settings__['mode_elas_deriv_min'] a = ScipyDerivative(self.__num_deriv_function__,input[met],order=self.__settings__['mode_elas_deriv_order'],dx=hstep,n=1,\ args=('v=self.' + self.__reactions__[react] + '()','self.' + self.__species__[met] + '=x')) except Exception, ex: print ex print '\nINFO: Elasticity evaluation failure in ', self.__reactions__[react], self.__species__[met] print 'Elasticity has been set to zero' print 'A stepsize that is too large might cause this ... try decreasing the factor and or min stepsize' print 'Keep in mind machine precision, is', self.__settings__['mach_floateps'], ' and if min stepsize' print 'becomes too small numeric error can become significant' a = 0.0 if abs(a) >= self.__settings__['mach_floateps']: if self.__settings__['display_debug'] == 1: print 'species: ' + self.__species__[met] print '--> d(' + self.__reactions__[react] + ')d(' + self.__species__[met] + ') = ' + str(a) self.__evmatrix__[react,met] = a # restore variables to ss values only if steady state used if self.__settings__['elas_evar_remap']: eval(compile(self.__remaps,'__remaps','exec')) #print "\nREMAPPING\n" self.elas_var_u = self.__evmatrix__ # If scaled mca is requested if self.__settings__['mode_mca_scaled'] == 1: Ds = numpy.zeros((len(input),len(input)),'d') Dj = numpy.zeros((len(input2),len(input2)),'d') for x in range(0,len(input)): Ds[x,x] = input[x] #print Ds for x in range(0,len(input2)): Dj[x,x] = 1.0/input2[x] if self.__settings__['elas_scaling_div0_fix'] and numpy.isinf(Dj[x,x]): print 'Infinite elasticity detected during scaling setting to zero (%s)' % self.reactions[x] Dj[x,x] = 1.0e-16 #print Dj Dj_e = numpy.dot(Dj,self.__evmatrix__) self.__evmatrix__ = numpy.dot(Dj_e, Ds) self.elas_var = self.__evmatrix__ else: self.elas_var = None if self.__settings__["elas_evar_upsymb"] == 1: # Upload elasticity symbols into namespace r,c = self.__evmatrix__.shape output2 = '' for x in range(0,r): react = self.__reactions__[x] for y in range (0,c): met = self.__D_s_Order[y] if self.__settings__['mode_mca_scaled']: setattr(self, 'ec' + react + '_' + met.replace('self.',''), self.__evmatrix__[x, y]) else: setattr(self, 'uec' + react + '_' + met.replace('self.',''), self.__evmatrix__[x, y]) # the new way of doing things (test phase) brett - 2007 self.ec = BagOfStuff(self.__evmatrix__, self.elas_var_row, self.elas_var_col) self.ec.load() if self.__settings__['mode_mca_scaled'] == 1: self.ec.scaled = True else: self.ec.scaled = False if self.__settings__['display_debug'] == 1: print '\n\n********************************\n' print output2 print '\n********************************\n' else: pass #print 'INFO: variable elasticity symbols not attached - .__settings__["elas_evar_upsymb"] = ' + `self.__settings__["elas_evar_upsymb"]` if self.__settings__['display_debug'] == 1: print '\ne_vmatrix' print `self.__D_s_Order`.replace('self.','') print self.__reactions__ print self.__evmatrix__
[docs] def CleanNaNsFromArray(self, arr, replace_val=0.0): """ Scan a matrix for NaN's and replace with zeros: - *arr* the array to be cleaned """ nantest = numpy.isnan(arr) if nantest.any(): for r in range(nantest.shape[0]): if nantest[r].any(): for c in range(nantest.shape[1]): if numpy.isnan(arr[r,c]): arr[r,c] = replace_val
[docs] def EvalEpar(self,input=None,input2=None): """ EvalEpar(input=None,input2=None) Calculate reaction elasticities towards the parameters. Both inputs (input1=species,input2=rates) should be valid (steady state for MCA) solutions and given in the correct order for them to be used. If either or both are missing the last state values are used automatically. Elasticities are scaled using input 1 and 2. Arguments: - input [default=None]: species concentration vector - input2 [default=None]: reaction rate vector Settings, set in mod.__settings__:: - elas_epar_upsymb [default = 1] attach individual elasticity symbols to model instance - elas_scaling_div0_fix [default=False] if NaN's are detected in the variable and parameter elasticity matrix replace with zero """ if input == None or input2 == None: input = self.state_species input2 = self.state_flux else: assert len(input) == len(self.species), 'length error this array must have ' + str(len(self.species)) + ' elements' assert len(input2) == len(self.reactions), 'length error this array must have ' + str(len(self.reactions)) + ' elements' # put in to fix the juicy bug - brett 20040930 # iow automatic derivatives use the current values of self.Sx to operate exec(self.__CDvarUpString__) # create parameter holding array parVal_hold = numpy.zeros((len(self.__parameters__)),'d') if self.__settings__['display_debug'] == 1: print '\nParinput = ' + `input` + '\n' print '\nparVal_hold1' print parVal_hold #Store parameter values into the storage array and copy them to the working array (parVal2) exec(self.__par_map2storeC) parVal2 = copy.copy(parVal_hold) # create the matrix of parameter elasticities self.__epmatrix__ = numpy.zeros((len(self.__reactions__),len(self.__parameters__)),'d') # create tuples of the rows and columns for the Ep matrix self.elas_par_row = tuple(copy.copy(self.__reactions__)) self.elas_par_col = tuple(self.__parameters__) for react in range(len(self.__reactions__)): if self.__settings__['display_debug'] == 1: print '\nReaction: ' + self.__reactions__[react] for par in range(len(self.__parameters__)): countV = 0 countPar = 0 try: """ I got the idea of scaling the stepsize to So from Herbert Sauro's Jarnac TModel.uEEOp function brett - 20040818""" hstep = getattr(self,self.__parameters__[par])*self.__settings__['mode_elas_deriv_factor'] # self.__settings__['mode_elas_deriv_factor'] = 0.0001 if abs(hstep) < self.__settings__['mode_elas_deriv_min']: # self.__settings__['mode_elas_deriv_min'] = 1.0e-12 hstep = self.__settings__['mode_elas_deriv_min'] a = ScipyDerivative(self.__num_deriv_function__,getattr(self,self.__parameters__[par]),order=self.__settings__['mode_elas_deriv_order'],dx=hstep,n=1,\ args=('v=self.' + self.__reactions__[react] + '()','self.' + self.__parameters__[par] + '=x')) except Exception, ex: print '\nNumeric derivative evaluation failure in ', self.__reactions__[react], self.__parameters__[par] print 'Elasticity has been set to NaN' print 'A stepsize that is too large might cause this ... try decreasing the factor and or min stepsize' print 'Keep in mind machine precision, is', self.__settings__['mach_floateps'], ' and if min stepsize' print 'becomes too small numeric error can become significant' print ex a = numpy.NaN if numpy.isnan(a) or abs(a) > self.__settings__['mach_floateps']: if self.__settings__['display_debug'] == 1: print 'parameter: ' + self.__parameters__[par] print '--> d(' + self.__reactions__[react] + ')d(' + self.__parameters__[par] + ') = ' + `a` self.__epmatrix__[react,par] = a self.elas_par_u = self.__epmatrix__ #Retrieve parameter values from the storage array exec(self.__par_remapC) if self.__settings__['display_debug'] == 1: print '\nparVal_hold2' print parVal_hold # Parameters are scaled by [1/J]*[Ep]*[P] # If scaled mca is requested scale self.__epmatrix__ if self.__settings__['mode_mca_scaled'] == 1: Dp = numpy.zeros((len(self.__parameters__),len(self.__parameters__)),'d') Dj = numpy.zeros((len(input2),len(input2)),'d') for x in range(0,len(self.__parameters__)): Dp[x,x] = parVal_hold[x] #print Dp for x in range(0,len(input2)): Dj[x,x] = 1.0/input2[x] if self.__settings__['elas_scaling_div0_fix'] and numpy.isinf(Dj[x,x]): print 'Infinite elasticity detected during scaling setting to zero (%s)' % self.reactions[x] Dj[x,x] = 1.0e-16 #print Dj Dj_e = numpy.dot(Dj,self.__epmatrix__) self.__epmatrix__ = numpy.dot(Dj_e, Dp) self.elas_par = self.__epmatrix__ else: self.elas_par = None del parVal_hold if self.__settings__["elas_epar_upsymb"] == 1: # Upload elasticity symbols into namespace r,c = self.__epmatrix__.shape output2 = '' for x in range(0,r): react = self.__reactions__[x] for y in range (0,c): met = self.__parameters__[y] if self.__settings__['mode_mca_scaled']: setattr(self, 'ec' + react + '_' + met.replace('self.',''), self.__epmatrix__[x, y]) else: setattr(self, 'uec' + react + '_' + met.replace('self.',''), self.__epmatrix__[x, y]) # the new way of doing things (test phase) brett - 2007 self.ecp = BagOfStuff(self.__epmatrix__, self.elas_par_row, self.elas_par_col) self.ecp.load() if self.__settings__['mode_mca_scaled'] == 1: self.ecp.scaled = True else: self.ecp.scaled = False if self.__settings__['display_debug'] == 1: print '\n\n********************************\n' print output2 print '\n********************************\n' if self.__settings__['display_debug'] == 1: print '\ne_pmatrix' print self.__parameters__ print self.__reactions__ print self.__epmatrix__
[docs] def showEvar(self,File=None): """ showEvar(File=None) Write out all variable elasticities as \'LaTeX\' formatted strings, alternatively write results to a file. Arguments: File [default=None]: an open writable Python file object """ # The Variable elasticities try: r,c = self.__evmatrix__.shape evar_output = '' for x in range(0,r): react = self.__reactions__[x] evar_output += '\n' + `self.__reactions__[x]` + '\n' for y in range (0,c): rtemp = self.__settings__['mode_number_format'] % self.__evmatrix__[x,y] met = self.__D_s_Order[y] if self.__evmatrix__[x,y] != 0.0: if self.__settings__['mode_mca_scaled']: elas = '\\ec{' + react + '}{' + met + '} = ' + rtemp else: elas = '\\uec{' + react + '}{' + met + '} = ' + rtemp evar_output += elas + '\n' evar_output = evar_output.replace('self.','') except: evar_output = 'No variable elasticities - run EvalEvar() to calculate' if File != None: print '\nspecies elasticities' File.write('\n## species elasticities\n') File.write(evar_output) else: print '\nspecies elasticities' print evar_output
[docs] def showEpar(self,File=None): """ showEpar(File=None) Write out all nonzero parameter elasticities as \'LaTeX\' formatted strings, alternatively write to file. Arguments: File [default=None]: an open writable Python file object """ # The parameter elasticities try: r,c = self.__epmatrix__.shape epar_output = '' for x in range(0,r): react = self.__reactions__[x] epar_output += '\n' + `self.__reactions__[x]` + '\n' for y in range (0,c): rtemp = self.__settings__['mode_number_format'] % self.__epmatrix__[x,y] met = self.__parameters__[y] # brett (paranoia removal) October 2004 #if abs(self.__epmatrix__[x,y]) >= 1.e-15: if self.__settings__['mode_mca_scaled']: elas = '\\ec{' + react + '}{' + met + '} = ' + rtemp else: elas = '\\uec{' + react + '}{' + met + '} = ' + rtemp #print elas epar_output += elas + '\n' epar_output = epar_output.replace('self.','') except: epar_output = 'No parameter elasticities - run EvalEpar() to calculate' if File != None: print '\nParameter elasticities' File.write('\n## Parameter elasticities\n') File.write(epar_output) else: print '\nParameter elasticities' print epar_output
[docs] def showElas(self,File=None): """ showElas(File=None) Print all elasticities to screen or file as \'LaTeX\' compatible strings. Calls showEvar() and showEpar() Arguments: File [default=None]: an open writable Python file object """ if File != None: self.showEvar(File) self.showEpar(File) else: self.showEvar() self.showEpar()
[docs] def ScaleKL(self,input,input2): """ ScaleKL(input,input2) Scale the K and L matrices with current steady state (if either input1 or 2 == None) or user input. Arguments: input: vector of species concentrations input2: vector of reaction rates """ # called by EvalEvar # Scale L matrix lmat = copy.copy(self.__lmatrix__) if type(input) == type(None) or type(input2) == type(None): input = self.state_species input2 = self.state_flux #print 'INFO: Using state_species and state_flux as input' else: assert len(input) == len(self.species), 'length error this array must have ' + str(len(self.species)) + ' elements' assert len(input2) == len(self.reactions), 'length error this array must have ' + str(len(self.reactions)) + ' elements' s_1_scale = numpy.zeros((len(input),len(input)),'d') for x in range(0,len(input)): try: s_1_scale[x,x] = 1.0/input[self.lmatrix_row[x]] #create 1/D Using the steady-state met's ordered to the rows except: print 'Zero species detected: ' + self.__species__[self.lmatrix_row[x]] + '_ss = ' + `input[self.lmatrix_row[x]]` s_1_scale[x,x] = 0.0 Si_order = numpy.zeros(len(self.lmatrix_col)) #check Si_scale = numpy.zeros((len(self.lmatrix_col),len(self.lmatrix_col)),'d') for x in range (0,len(self.lmatrix_col)): Si_order[x] = self.lmatrix_col[x] #check Si_scale[x,x] = input[self.lmatrix_col[x]] L_Si = numpy.dot(lmat, Si_scale) L_scaled = numpy.dot(s_1_scale, L_Si) self.lmatrix_scaled = L_scaled #Scale K matrix kmat = copy.copy(self.__kmatrix__) j_1_scale = numpy.zeros((len(input2),len(input2)),'d') for x in range(0,len(input2)): try: j_1_scale[x,x] = 1.0/input2[self.kmatrix_row[x]] except: print '\nNull flux detected: ' + self.__reactions__[self.kmatrix_row[x]] + '_ss = ' + `input2[self.kmatrix_row[x]]` j_1_scale[x,x] = 0.0 Ji_order = numpy.zeros(len(self.kmatrix_col)) #check Ji_scale = numpy.zeros((len(self.kmatrix_col),len(self.kmatrix_col)),'d') for x in range (0,len(self.kmatrix_col)): Ji_order[x] = self.kmatrix_col[x] #check Ji_scale[x,x] = input2[self.kmatrix_col[x]] K_Ji = numpy.dot(kmat, Ji_scale) K_scaled = numpy.dot(j_1_scale, K_Ji) self.kmatrix_scaled = K_scaled
def __num_deriv_function__(self,x,react,met): """ __num_deriv_function__(x,react,met) System function that evaluates the rate equation, used by numeric perturbation methods to derive elasticities. It uses a specific format assigning x to met and evaluating react for v and is tailored for the ScipyDerivative() function using precompiled function strings. Arguments: x: value to assign to met react: reaction met: species """ exec(met) self.Forcing_Function() exec(react) return v # Control coefficient routines
[docs] def EvalCC(self): """ EvalCC() Calculate the MCA control coefficients using the current steady-state solution. mod.__settings__["mca_ccj_upsymb"] = 1 attach the flux control coefficients to the model instance mod.__settings__["mca_ccs_upsymb"] = 1 attach the concentration control coefficients to the model instance Arguments: None """ #sort E to match K and L e2 = numpy.zeros((len(self.kmatrix_row), len(self.lmatrix_row)), 'd') #print e2 for x in range (0,len(self.kmatrix_row)): for y in range (0,len(self.lmatrix_row)): e2[x,y] = self.elas_var_u[self.kmatrix_row[x],self.lmatrix_row[y]] if self.__settings__['display_debug'] == 1: print self.lmatrix_row print self.kmatrix_row print '---' print self.__nmatrix__.shape print self.nmatrix_row print self.nmatrix_col print self.nmatrix print '---' print self.__lmatrix__.shape print self.__lmatrix__ print '---' print self.__kmatrix__.shape print self.__kmatrix__ print '---' print e2.shape print e2 EL = -1.0*numpy.dot(e2,self.__lmatrix__) KEL = numpy.concatenate((self.__kmatrix__, EL),1) self.kel_unscaled = KEL go = 0 try: Ci_u = scipy.linalg.inv(KEL) # fortran has a very fp idea of zero I use abs(val)<1.0e-15 self.__structural__.MatrixFloatFix(Ci_u,val=1.e-15) go = 1 except Exception, ex: print ex print '\nINFO: K-EL matrix inversion failed this is possibly due to NaN values in the Elasticity matrix' print 'NaN elasticities can be caused by zero fluxes or concentrations look at the settings (in mod.__settings__)' print '\'elas_scaling_div0_fix\' and \'elas_zero_flux_fix\'' go = 0 if go == 1 and not self.__settings__['mode_mca_scaled']: self.mca_ci = Ci_u self.__mca_CCi_unscaled = Ci_u del Ci_u elif go == 1 and self.__settings__['mode_mca_scaled']: Vi_l = self.__kmatrix__.shape[1] #Ind fluxes Vd_l = abs(self.__kmatrix__.shape[0] - self.__kmatrix__.shape[1]) # Can never be the case but b.paranoid Si_l = self.__lmatrix__.shape[1] #Ind species Sd_l = abs(self.__lmatrix__.shape[0] - self.__lmatrix__.shape[1]) # Can never be the case but b.paranoid scale1_l = Vi_l + Si_l scale1 = numpy.zeros((scale1_l,scale1_l),'d') for x in range(0,Vi_l): scale1[x,x] = 1.0/self.state_flux[self.kmatrix_row[x]] for x in range(Vi_l,scale1_l): scale1[x,x] = 1.0/self.state_species[self.lmatrix_row[x - Vi_l]] scale2 = numpy.zeros((Vi_l + Vd_l, Vi_l + Vd_l),'d') for x in range(0,len(self.state_flux)): scale2[x,x] = self.state_flux[self.kmatrix_row[x]] left = numpy.dot(scale1, Ci_u) Ci = numpy.dot(left, scale2) self.mca_ci = Ci del scale1_l, scale1, scale2, left, Ci, Ci_u Cirow = [] Cicol = [] # the wierdness continues for x in range (0,len(self.kmatrix_row)): Cicol.append(self.__reactions__[self.kmatrix_row[x]]) for x in range (0,len(self.kmatrix_col)): Cirow.append(self.__reactions__[self.kmatrix_col[x]]) for x in range (0,len(self.lmatrix_col)): Cirow.append(self.__D_s_Order[self.lmatrix_col[x]].replace('self.','')) self.mca_ci_row = Cirow self.mca_ci_col = Cicol del e2, EL, KEL if self.__settings__['display_debug'] == 1: print 'print self.mca_ci_row' print self.mca_ci_row print 'print self.mca_ci_col' print self.mca_ci_col print 'print self.mca_ci' print self.mca_ci CJi = numpy.zeros((len(self.kmatrix_col),self.mca_ci.shape[1]),'d') CSi = numpy.zeros((self.mca_ci.shape[0] - len(self.kmatrix_col),self.mca_ci.shape[1]),'d') for x in range (0, self.mca_ci.shape[0]): for y in range (0, self.mca_ci.shape[1]): if x < len(self.kmatrix_col): CJi[x,y] = self.mca_ci[x,y] else: CSi[x - len(self.kmatrix_col),y] = self.mca_ci[x,y] Ko_row = [] Ko_col = [] sKo = numpy.zeros(self.__kzeromatrix__.shape,'d') xFactor = self.__kmatrix__.shape[0] - self.__kzeromatrix__.shape[0] #we need the scaled k/l matrices to calculate the dependent cc's #self.ScaleKL() for x in range (self.__kzeromatrix__.shape[0]-1,-1,-1): Ko_row.append(self.__reactions__[self.kzeromatrix_row[x]]) for y in range (self.__kzeromatrix__.shape[1]-1,-1,-1): #this can be replaced by a row slice operation sKo[x,y] = self.kmatrix_scaled[x+xFactor, y] if x == 0: Ko_col.append(self.__reactions__[self.kzeromatrix_col[y]]) Ko_row.reverse() Ko_col.reverse() if self.__settings__['display_debug'] == 1: print 'CJi' print CJi print 'sKo' print sKo print 'self.kzeromatrix' print self.__kzeromatrix__ #new if self.__settings__['mode_mca_scaled']: self.mca_cjd = numpy.dot(sKo, CJi) else: self.mca_cjd = numpy.dot(self.__kzeromatrix__, CJi) self.mca_cjd_row = Ko_row self.mca_cjd_col = copy.copy(self.mca_ci_col) del sKo, CJi if self.__settings__['display_debug'] == 1: print 'self.mca_cjd_row' print self.mca_cjd_row print 'self.mca_cjd_col' print self.mca_cjd_col print 'self.mca_cjd' print self.mca_cjd Lo_row = [] Lo_col = [] sLo = numpy.zeros(self.__lzeromatrix__.shape,'d') xFactor = self.__lmatrix__.shape[0] - self.__lzeromatrix__.shape[0] for x in range (self.__lzeromatrix__.shape[0]-1,-1,-1): Lo_row.append(self.__D_s_Order[self.lzeromatrix_row[x]].replace('self.','')) for y in range (self.__lzeromatrix__.shape[1]-1,-1,-1): #this can be replaced by a row slice operation sLo[x,y] = self.lmatrix_scaled[x+xFactor,y] if x == 0: Lo_col.append(self.__D_s_Order[self.lzeromatrix_col[y]]) Lo_row.reverse() Lo_col.reverse() if self.__HAS_MOIETY_CONSERVATION__ == True: #Only do this if there is S dependency if self.__settings__['mode_mca_scaled']: self.mca_csd = numpy.dot(sLo, CSi) else: self.mca_csd = numpy.dot(self.__lzeromatrix__, CSi) self.mca_csd_row = Lo_row self.mca_csd_col = copy.copy(self.mca_ci_col) else: self.mca_csd = None self.mca_csd_row = None self.mca_csd_col = None del CSi, sLo if self.__settings__['display_debug'] == 1: print 'self.mca_csd' print self.mca_csd print 'self.mca_csd_row' print self.mca_csd_row print 'self.mca_csd_col' print self.mca_csd_col if self.__HAS_FLUX_CONSERVATION__ and self.__HAS_MOIETY_CONSERVATION__: self.cc_flux = numpy.concatenate((self.mca_ci[:self.__kmatrix__.shape[1],:],self.mca_cjd)) self.cc_flux_row = copy.copy(self.mca_ci_row[:self.__kmatrix__.shape[1]] + self.mca_cjd_row) self.cc_flux_col = copy.copy(self.mca_ci_col) self.cc_conc = numpy.concatenate((self.mca_ci[self.__kmatrix__.shape[1]:,:],self.mca_csd)) self.cc_conc_row = copy.copy(self.mca_ci_row[self.__kmatrix__.shape[1]:] + self.mca_csd_row) self.cc_conc_col = copy.copy(self.mca_ci_col) elif self.__HAS_FLUX_CONSERVATION__: self.cc_flux = numpy.concatenate((self.mca_ci[:self.__kmatrix__.shape[1],:],self.mca_cjd)) self.cc_flux_row = copy.copy(self.mca_ci_row[:self.__kmatrix__.shape[1]] + self.mca_cjd_row) self.cc_flux_col = copy.copy(self.mca_ci_col) self.cc_conc = copy.copy(self.mca_ci[self.__kmatrix__.shape[1]:,:]) self.cc_conc_row = copy.copy(self.mca_ci_row[self.__kmatrix__.shape[1]:]) self.cc_conc_col = copy.copy(self.mca_ci_col) elif self.__HAS_MOIETY_CONSERVATION__: print 'INFO: this is interesting no dependent flux cc\'s only dependent conc cc\'s!' self.cc_flux = copy.copy(self.mca_ci[:self.__kmatrix__.shape[1],:]) self.cc_flux_row = copy.copy(self.mca_ci_row[:self.__kmatrix__.shape[1]]) self.cc_flux_col = copy.copy(self.mca_ci_col) self.cc_conc = numpy.concatenate((self.mca_ci[self.__kmatrix__.shape[1]:,:],self.mca_csd)) self.cc_conc_row = copy.copy(self.mca_ci_row[self.__kmatrix__.shape[1]:] + self.mca_csd_row) self.cc_conc_col = copy.copy(self.mca_ci_col) else: print 'INFO: this is interesting no dependent flux/conc coefficients!' self.cc_flux = copy.copy(self.mca_ci[:self.__kmatrix__.shape[1],:]) self.cc_flux_row = copy.copy(self.mca_ci_row[:self.__kmatrix__.shape[1]]) self.cc_flux_col = copy.copy(self.mca_ci_col) self.cc_conc = copy.copy(self.mca_ci[self.__kmatrix__.shape[1]:,:]) self.cc_conc_row = copy.copy(self.mca_ci_row[self.__kmatrix__.shape[1]:]) self.cc_conc_col = copy.copy(self.mca_ci_col) self.cc_all = numpy.concatenate((self.cc_flux,self.cc_conc)) self.cc_all_row = self.cc_flux_row + self.cc_conc_row self.cc_all_col = copy.copy(self.mca_ci_col) # the new way of doing things (test phase) brett - 2007 self.cc = BagOfStuff(self.cc_all, self.cc_all_row, self.cc_all_col) self.cc.load() if self.__settings__['mode_mca_scaled'] == 1: self.cc.scaled = True else: self.cc.scaled = False if self.__settings__['display_debug'] == 1: print 'self.cc_flux' print self.cc_flux_row print self.cc_flux_col print self.cc_flux.shape print 'self.cc_conc' print self.cc_conc_row print self.cc_conc_col print self.cc_conc.shape print 'self.cc_all' print self.cc_all_row print self.cc_all_col print self.cc_all.shape if self.__settings__["mca_ccj_upsymb"]: #CJ r,c = self.cc_all.shape CJoutput = '' for x in range(0,len(self.__reactions__)): for y in range (0,c): if self.__settings__['mode_mca_scaled']: setattr(self, 'ccJ' + self.cc_all_row[x] + '_' + self.cc_all_col[y], self.cc_all[x,y]) else: setattr(self, 'uccJ' + self.cc_all_row[x] + '_' + self.cc_all_col[y], self.cc_all[x,y]) if self.__settings__["mca_ccs_upsymb"]: #CS CSoutput = '' for x in range(len(self.__reactions__),r): for y in range (0,c): if self.__settings__['mode_mca_scaled']: setattr(self, 'cc' + self.cc_all_row[x] + '_' + self.cc_all_col[y], self.cc_all[x,y]) else: setattr(self, 'ucc' + self.cc_all_row[x] + '_' + self.cc_all_col[y], self.cc_all[x,y]) if self.__settings__['display_debug'] == 1: print 'CJoutput' print CJoutput print 'CSoutput' print CSoutput
[docs] def showCC(self,File=None): """ showCC(File=None) Print all control coefficients as \'LaTex\' formatted strings to the screen or file. Arguments: File [default=None]: an open, writable Python file object """ r,c = self.cc_all.shape if self.__settings__["mca_ccall_altout"]: CAltoutput = '' for x in range(0,c): col = self.cc_all_col[x] CAltoutput += '\n`Reaction ' + col + '\'\n' for y in range(0,r): if y == 0: CAltoutput += '\"Flux control coefficients\"\n' if y == len(self.__reactions__): CAltoutput += '\"Concentration control coefficients\"\n' row = self.cc_all_row[y] rtemp = self.__settings__['mode_number_format'] % self.cc_all[y,x] if y < len(self.__reactions__): if self.__settings__['mode_mca_scaled']: cc = '\\cc{J' + row + '}{' + col + '} = ' + rtemp else: cc = '\\ucc{J' + row + '}{' + col + '} = ' + rtemp else: if self.__settings__['mode_mca_scaled']: cc = '\\cc{' + row + '}{' + col + '} = ' + rtemp else: cc = '\\ucc{' + row + '}{' + col + '} = ' + rtemp CAltoutput += cc + '\n' else: if self.__settings__["mca_ccall_fluxout"]: #CJ CJoutput = '' for x in range(0,len(self.__reactions__)): row = self.cc_all_row[x] CJoutput += '\n`J' + row + '\'\n' for y in range (0,c): col = self.cc_all_col[y] rtemp = self.__settings__['mode_number_format'] % self.cc_all[x,y] if self.__settings__['mode_mca_scaled']: cc = '\\cc{J' + row + '}{' + col + '} = ' + rtemp else: cc = '\\ucc{J' + row + '}{' + col + '} = ' + rtemp CJoutput += cc + '\n' if self.__settings__["mca_ccall_concout"]: #CS CSoutput = '' for x in range(len(self.__reactions__),r): row = self.cc_all_row[x] CSoutput += '\n`' + row + '\'\n' for y in range (0,c): col = self.cc_all_col[y] rtemp = self.__settings__['mode_number_format'] % self.cc_all[x,y] if self.__settings__['mode_mca_scaled']: cc = '\\cc{' + row + '}{' + col + '} = ' + rtemp else: cc = '\\ucc{' + row + '}{' + col + '} = ' + rtemp CSoutput += cc + '\n' if File != None: #assert type(File) == file, 'showCC() needs an open file object' if self.__settings__["mca_ccall_altout"]: print '\nControl coefficients grouped by reaction' File.write('\n## Control coefficients grouped by reaction\n') File.write(CAltoutput) else: if self.__settings__["mca_ccall_fluxout"]: print '\nFlux control coefficients' File.write('\n## Flux control coefficients\n') File.write(CJoutput) if self.__settings__["mca_ccall_concout"]: print '\nConcentration control coefficients' File.write('\n## Concentration control coefficients\n') File.write(CSoutput) else: if self.__settings__["mca_ccall_altout"]: print '\nControl coefficients grouped by reaction' print CAltoutput else: if self.__settings__["mca_ccall_fluxout"]: print '\nFlux control coefficients' print CJoutput if self.__settings__["mca_ccall_concout"]: print '\nConcentration control coefficients' print CSoutput
[docs] def EvalRC(self): """ EvalRC() Calculate the MCA response coefficients using the current steady-state solution. Arguments: None """ reordered_cc_all = numpy.zeros(self.cc_all.shape,'d') for reac in range(len(self.elas_par_row)): reordered_cc_all[:,reac] = self.cc_all[:,list(self.cc_all_col).index(self.elas_par_row[reac])] self.mca_rc_par = numpy.dot(reordered_cc_all,self.elas_par) self.mca_rc_par_row = self.cc_all_row self.mca_rc_par_col = self.elas_par_col del reordered_cc_all self.__structural__.MatrixFloatFix(self.mca_rc_par) self.rc = BagOfStuff(self.mca_rc_par, self.mca_rc_par_row, self.mca_rc_par_col) self.rc.load() if self.__settings__['mode_mca_scaled'] == 1: self.rc.scaled = True else: self.rc.scaled = False
[docs] def EvalRCT(self): """ EvalRCT() Calculate the MCA response coefficients using the current steady-state solution. Responses to changes in the sums of moiety conserved cycles are also calculated. Arguments: None """ # We arbitrarily choose the order of reactions as that of elas_var_row # and the order of species as that of elas_var_col. The order of dependent # species (there is one for each moiety conserved cycle) is from the # conservation matrix (mod.Consmatrix.row). The final output is the # (m x (m-r)) R^S_T matrix and the (n x (m-r)) R^J_T matrix. # # See "Metabolic control analysis in a nutshell" by Hofmeyr and also # Kholodenko, Sauro, Westerhoff 1994 for the matrix formulations. # # Danie Palm - 2011-10-26 # Reorder reactions in Cs and CJ to match elas_var_row reaction_reordered_cc_conc = numpy.zeros(self.cc_conc.shape,'d') for reac in range(len(self.elas_var_row)): reaction_reordered_cc_conc[:,reac] = self.cc_conc[:,list(self.cc_conc_col).index(self.elas_var_row[reac])] reaction_reordered_cc_flux = numpy.zeros(self.cc_flux.shape,'d') for reac in range(len(self.elas_var_row)): reaction_reordered_cc_flux[:,reac] = self.cc_flux[:,list(self.cc_flux_col).index(self.elas_var_row[reac])] # Reorder metabolites in Cs to match elas_var_col reordered_cc_conc = numpy.zeros(self.cc_conc.shape,'d') for metab in range(len(self.elas_var_col)): reordered_cc_conc[metab,:] = reaction_reordered_cc_conc[list(self.cc_conc_row).index(self.elas_var_col[metab]),:] # Reorder fluxes in CJ to match elas_var_row reordered_cc_flux = numpy.zeros(self.cc_flux.shape,'d') for flux in range(len(self.elas_var_row)): reordered_cc_flux[flux,:] = reaction_reordered_cc_flux[list(self.cc_flux_row).index(self.elas_var_row[flux]),:] # Some basic dimensions and matrices m = len(self.species) r = len(self.Lmatrix.col) Im = numpy.matrix(numpy.eye(m)) # Construct the (right) pseudoinverse of the the conservation matrix in the order of elas_var_col reordered_zero_I = numpy.zeros((m, m-r), 'd') for species in self.elas_var_col: if species in self.Consmatrix.row: row = self.elas_var_col.index(species) col = self.Consmatrix.row.index(species) reordered_zero_I[row,col] = 1.0 # Unlike normal RCs, separate calculations are required for scaled # and unscaled RCT's. We also need to pick the right elasticity # matrix: elas_var for scaled and elas_var_u for unscaled. if self.__settings__['mode_mca_scaled'] == 1: # Some scaling matrices diag_T = None if self.__KeyWords__['Species_In_Conc']: diag_T = numpy.matrix(numpy.diag(self.__tvec_c__)) else: diag_T = numpy.matrix(numpy.diag(self.__tvec_a__)) diag_S = numpy.matrix(numpy.diag(self.data_sstate.getStateData(*self.elas_var_col))) # Calculate concentration responses to changes in T # (Cs*es + Im) * diag_S.I * reordered_zero_I * diag_T self.mca_rct_conc = numpy.dot(numpy.dot(numpy.dot(numpy.dot(reordered_cc_conc, self.elas_var) + Im, numpy.linalg.inv(diag_S)), reordered_zero_I), diag_T) # Calculate flux responses to changes in T # CJ*es * diag_S.I * reordered_zero_I * diag_T self.mca_rct_flux = numpy.dot(numpy.dot(numpy.dot(numpy.dot(reordered_cc_flux, self.elas_var), numpy.linalg.inv(diag_S)), reordered_zero_I), diag_T) else: # Calculate concentration responses to changes in T # (Cs*es + Im) * reordered_zero_I self.mca_rct_conc = numpy.dot(numpy.dot(reordered_cc_conc, self.elas_var_u) + Im, reordered_zero_I) # Calculate flux responses to changes in T # CJ*es * reordered_zero_I self.mca_rct_flux = numpy.dot(numpy.dot(reordered_cc_flux, self.elas_var_u), reordered_zero_I) # Cleanup del reordered_cc_conc del reordered_cc_flux # We simply prepend 'T_' to the name of the dependent species to prevent # confusion with the real species. This is a parameter, not variable. moiety_sum_names = ['T_{0}'.format(dependent_species) for dependent_species in self.Consmatrix.row] self.__structural__.MatrixFloatFix(self.mca_rct_conc) self.mca_rct_conc_row = self.elas_var_col self.mca_rct_conc_col = moiety_sum_names self.__structural__.MatrixFloatFix(self.mca_rct_flux) self.mca_rct_flux_row = self.elas_var_row self.mca_rct_flux_col = moiety_sum_names # Augment the existing m.rc object (this could be more efficient) # Reordering as a matter of principle: too many burnt fingers vstacked_col = list(self.elas_var_row) + list(self.elas_var_col) vstacked = numpy.vstack((self.mca_rct_flux, self.mca_rct_conc)) reordered_vstack = numpy.zeros(vstacked.shape, 'd') for row_index in range(len(self.rc.row)): reordered_vstack[row_index,:] = vstacked[list(vstacked_col).index(self.rc.row[row_index]),:] hstacked = numpy.hstack((self.rc.matrix, vstacked)) self.rc = BagOfStuff(hstacked, self.rc.row, self.rc.col + moiety_sum_names) self.rc.load() if self.__settings__['mode_mca_scaled'] == 1: self.rc.scaled = True else: self.rc.scaled = False
[docs] def EvalEigen(self): """ EvalEigen() Calculate the eigenvalues or vectors of the unscaled Jacobian matrix and thereby analyse the stability of a system Arguments: None """ Nr = copy.copy(self.__nrmatrix__) #sort E to match K and L Es = numpy.zeros((self.elas_var_u.shape), 'd') #print e2 for x in range (0,len(self.nmatrix_col)): for y in range (0,len(self.lmatrix_row)): Es[x,y] = self.elas_var_u[self.nmatrix_col[x],self.lmatrix_row[y]] if self.__HAS_MOIETY_CONSERVATION__ == True: jac1 = numpy.dot(Nr,Es) jacobian = numpy.dot(jac1,copy.copy(self.__lmatrix__)) else: jacobian = numpy.dot(Nr,Es) Si = [] for x in range(0,len(self.lmatrix_col)): Si.append(self.__species__[self.lmatrix_col[x]]) self.jacobian = tuple(jacobian) self.jacobian_row = tuple(Si) self.jacobian_col = tuple(Si) if self.__settings__['mode_eigen_output']: #returns eigenvalues as well as left and right eigenvectors eigenval,self.eigen_vecleft,self.eigen_vecright = scipy.linalg.eig(jacobian,numpy.identity(jacobian.shape[0],'d'),left=1,right=1) if self.__settings__['display_debug'] == 1: print '\nEigenvalues' print eigenval print '\nLeft Eigenvector' print self.eigen_vecleft print '\nRight Eigenvector' print self.eigen_vecright else: #returns eigenvalues eigenval = scipy.linalg.eigvals(jacobian) if self.__settings__['display_debug'] == 1: print '\nEigenvalues' print eigenval self.eigen_values = eigenval #self.eigen_order = tuple(eigenorder) for x in range(len(self.eigen_values)): setattr(self, 'lambda' + str(x+1), self.eigen_values[x]) if self.__settings__['display_debug'] == 1: print '\nEigenvalues attached as lambda1 ... lambda' + `len(eigenorder)` + '\n'
[docs] def showEigen(self,File=None): """ showEigen(File=None) Print the eigenvalues and stability analysis of a system generated with EvalEigen() to the screen or file. Arguments: File [default=None]: an open, writable Python file object """ eigenstats = '' eigenstats += 'Max real part: ' + self.__settings__['mode_number_format'] % max(self.eigen_values.real) + '\n' eigenstats += 'Min real part: ' + self.__settings__['mode_number_format'] % min(self.eigen_values.real) + '\n' eigenstats += 'Max absolute imaginary part: ' + self.__settings__['mode_number_format'] % max(abs(self.eigen_values.imag)) + '\n' eigenstats += 'Min absolute imaginary part: ' + self.__settings__['mode_number_format'] % min(abs(self.eigen_values.imag)) + '\n' eigenstats += 'Stiffness: ' + self.__settings__['mode_number_format'] % (max(abs(self.eigen_values.real))/min(abs(self.eigen_values.real))) + '\n' pure_real = 0 pure_imag = 0 negcplx = 0 pure_cplx = 0 pure_zero = 0 pos_real = 0 neg_real = 0 for x in self.eigen_values: if x.real != 0.0 and x.imag == 0.0: pure_real += 1 if x.real == 0.0 and x.imag != 0.0: pure_imag += 1 if x.real == 0.0 and x.imag == 0.0: pure_zero += 1 if x.real > 0.0: pos_real += 1 if x.real < 0.0: neg_real += 1 if x.real < 0.0 and x.imag != 0.0: negcplx += 1 eigenstats += '\n## Stability' eiglen = len(self.eigen_values) if neg_real == eiglen: eigenstats += ' --> Stable state' elif pure_zero > 0: eigenstats += ' --> Undetermined' elif pos_real == eiglen: eigenstats += ' --> Unstable state' eigenstats += '\nPurely real: ' + `pure_real` eigenstats += '\nPurely imaginary: ' + `pure_imag` eigenstats += '\nZero: ' + `pure_zero` eigenstats += '\nPositive real part: ' + `pos_real` eigenstats += '\nNegative real part: ' + `neg_real` if File != None: #assert type(File) == file, 'showEigen() needs an open file object' print '\nEigen values' File.write('\n## Eigen values\n') scipy.io.write_array(File,self.eigen_values,precision=2,keep_open=1) print '\nEigen statistics' File.write('\n## Eigen statistics\n') File.write(eigenstats) if self.__settings__['mode_eigen_output']: print '\nLeft eigen vector' File.write('\n## Left eigen vector\n') scipy.io.write_array(File,self.eigen_vecleft,precision=2,keep_open=1) print '\nRight eigen vector' File.write('\n## Right eigen vector\n') scipy.io.write_array(File,self.eigen_vecright,precision=2,keep_open=1) else: print '\nEigen values' print self.eigen_values print '\nEigen statistics' print eigenstats if self.__settings__['mode_eigen_output']: print '\nLeft eigen vector' print self.eigen_vecleft print '\nRight eigen vector' print self.eigen_vecright #Utility functions #new generation metafunctions ## def doLoad(self,stoich_load=0): ## """ ## doLoad(stoich_load=0) ## Load and instantiate a PySCeS model so that it can be used for further analyses. ## Calls model loading subroutines: ## Stoichiometry_Analyse() [override=0,load=stoich_load] ## InitialiseModel() ## Arguments: ## stoich_load [default=0]: try to load a stoichiometry saved with Stoichiometry_Save_Serial() ## """ ## self.InitialiseInputFile() ## assert self.__parseOK, '\nError in input file, parsing could not complete' ## self.Stoichiometry_Analyse(override=0,load=stoich_load) ## # add new Style functions to model ## self.InitialiseFunctions() ## self.InitialiseCompartments() ## self.InitialiseRules() ## self.InitialiseEvents() ## self.InitialiseOldFunctions() # TODO replace this with initialisation functions ## self.InitialiseModel() ## self.InitialiseRuleChecks()
[docs] def doSim(self,end=10.0,points=21): """ doSim(end=10.0,points=20.0) Run a time simulation from t=0 to t=sim_end with sim_points. Calls: Simulate() Arguments: end [default=10.0]: simulation end time points [default=20.0]: number of points in the simulation """ self.sim_end = end self.sim_points = points self.Simulate()
[docs] def doSimPlot(self, end=10.0, points=21, plot='species', fmt='lines', filename=None): """ Run a time simulation from t=0 to t=sim_end with sim_points and plot the results. The required output data and format can be set: - *end* the end time (default=10.0) - *points* the number of points in the simulation (default=20.0) - *plot* (default='species') select output data - 'species' - 'rates' - 'all' both species and rates - *fmt* plot format, UPI backend dependent (default='') or the *CommonStyle* 'lines' or 'points'. - *filename* if not None (default) then the plot is exported as *filename*.png Calls: - **Simulate()** - **SimPlot()** """ self.sim_end = end self.sim_points = points self.Simulate() self.SimPlot(plot=plot, format=fmt, filename=filename)
[docs] def exportSimAsSedML(self, output='files', return_sed=False, vc_given='PySCeS', vc_family='Software', vc_email='', vc_org='pysces.sourceforge.net'): """ Exports the current simulation as SED-ML in various ways it creates and stores the SED-ML files in a folder generated from the model name. - *output* [default='files'] the SED-ML export type can be one or more comma separated e.g. 'files,combine' - *files* export the plain SBML and SEDML XML files - *archive* export as a SED-ML archive *<file>.sedx* containing the SBML and SEDML xml files - *combine* export as a COMBINE archive *<file>.omex* containing the SBML, SEDML, manifest (XML) and metadata (RDF) - *vc_given* [default='PySCeS'] - *vc_family* [default='Software'] - *vc_email* [default='bgoli@users.sourceforge.net'] - *vc_org* [default='<pysces.sourceforge.net>'] """ sedname = self.ModelFile.replace('.psc','') sedout = os.path.join(OUTPUT_DIR, 'sedout') if not os.path.exists(sedout): os.makedirs(sedout) S = SED.SED(sedname+'_sed', sedout) S.addModel(sedname, self) S.addSimulation('sim0', self.sim_start, self.sim_end, self.sim_points,\ self.__SIMPLOT_OUT__, initial=None, algorithm='KISAO:0000019') S.addTask('task0', 'sim0', sedname) S.addTaskDataGenerators('task0') S.addTaskPlot('task0') output = [s_.strip() for s_ in output.split(',')] for s_ in output: if s_ == 'files': S.writeSedScript() S.writeSedXML() elif s_ == 'archive': S.writeSedXArchive() elif s_ == 'combine': S.writeCOMBINEArchive(vc_given=vc_given, vc_family=vc_family, vc_email=vc_email, vc_org=vc_org) S._SED_CURRENT_ = False if return_sed: return S else: del S
[docs] def doSimPerturb(self,pl,end): """ **Deprecated**: use events instead """ pass
[docs] def doState(self): """ doState() Calculate the steady-state solution of the system. Calls: State() Arguments: None """ self.State()
[docs] def doStateShow(self): """ doStateShow() Calculate the steady-state solution of a system and show the results. Calls: State() showState() Arguments: None """ self.State() assert self.__StateOK__ == True, '\n\nINFO: Invalid steady state: run mod.doState() and check for errors' self.showState()
[docs] def doElas(self): """ doElas() Calculate the model elasticities, this method automatically calculates a steady state. Calls: State() EvalEvar() EvalEpar() Arguments: None """ self.State() assert self.__StateOK__ == True, '\n\nINFO: Invalid steady state: run mod.doState() and check for errors' self.EvalEvar() self.EvalEpar()
[docs] def doEigen(self): """ doEigen() Calculate the eigenvalues, automatically performs a steady state and elasticity analysis. Calls: State() EvalEvar() Evaleigen() Arguments: None """ self.State() assert self.__StateOK__ == True, '\n\nINFO: Invalid steady state: run mod.doState() and check for errors' self.__settings__["elas_evar_upsymb"] = 1 self.EvalEvar() self.__settings__["elas_evar_upsymb"] = 1 self.EvalEigen()
[docs] def doEigenShow(self): """ doEigenShow() Calculate the eigenvalues, automatically performs a steady state and elasticity analysis and displays the results. Calls: doEigen() showEigen() Arguments: None """ self.doEigen() self.showEigen()
[docs] def doEigenMca(self): """ doEigenMca() Calculate a full Control Analysis and eigenvalues, automatically performs a steady state, elasticity, control analysis. Calls: State() EvalEvar() EvalCC() Evaleigen() Arguments: None """ self.State() assert self.__StateOK__ == True, '\n\nINFO: Invalid steady state: run mod.doState() and check for errors' self.EvalEvar() self.EvalCC() self.EvalEigen()
[docs] def doMca(self): """ doMca() Perform a complete Metabolic Control Analysis on the model, automatically calculates a steady state. Calls: State() EvalEvar() EvalEpar() EvalCC() Arguments: None """ self.State() assert self.__StateOK__ == True, '\n\nINFO: Invalid steady state run: mod.doState() and check for errors' self.EvalEvar() self.EvalEpar() self.EvalCC()
[docs] def doMcaRC(self): """ doMca() Perform a complete Metabolic Control Analysis on the model, automatically calculates a steady state. Calls: State() EvalEvar() EvalEpar() EvalCC() EvalRC() Arguments: None """ self.State() assert self.__StateOK__ == True, '\n\nINFO: Invalid steady state run: mod.doState() and check for errors' self.EvalEvar() self.EvalEpar() self.EvalCC() self.EvalRC()
[docs] def doMcaRCT(self): """ doMcaRCT() Perform a complete Metabolic Control Analysis on the model, automatically calculates a steady state. In additional, response coefficients to the sums of moiety-conserved cycles are calculated. Calls: State() EvalEvar() EvalEpar() EvalCC() EvalRC() EvalRCT() Arguments: None """ self.State() assert self.__StateOK__ == True, '\n\nINFO: Invalid steady state run: mod.doState() and check for errors' self.EvalEvar() self.EvalEpar() self.EvalCC() self.EvalRC() if self.__HAS_MOIETY_CONSERVATION__: self.EvalRCT() else: print 'No moiety conservation detected, reverting to simple doMcaRC().' #show/save function prototypes
[docs] def showSpecies(self,File=None): """ showSpecies(File=None) Prints the current value of the model's variable species (mod.X) to screen or file. Arguments: File [default=None]: an open, writable Python file object """ out_list = [] print '\nSpecies values' out_list.append('\n## species values\n') for x in range(len(self.__species__)): if File == None: ## print self.__species__[x] + ' = ' + self.__settings__['mode_number_format'] % eval('self.' + self.__species__[x]) print self.__species__[x] + ' = ' + self.__settings__['mode_number_format'] % getattr(self, self.__species__[x]) else: ## out_list.append(self.__species__[x] + ' = ' + self.__settings__['mode_number_format'] % eval('self.' + self.__species__[x]) + '\n') out_list.append(self.__species__[x] + ' = ' + self.__settings__['mode_number_format'] % getattr(self, self.__species__[x]) + '\n') if File != None: for x in out_list: File.write(x)
[docs] def showSpeciesI(self,File=None): """ showSpeciesI(File=None) Prints the current value of the model's variable species initial values (mod.X_init) to screen or file. Arguments: File [default=None]: an open, writable Python file object """ out_list = [] print '\nSpecies initial values' out_list.append('\n## species initial values\n') for x in range(len(self.__species__)): if File == None: print self.__species__[x] + '_init = ' + self.__settings__['mode_number_format'] % getattr(self, self.__species__[x]+'_init') else: out_list.append(self.__species__[x] + ' = ' + self.__settings__['mode_number_format'] % getattr(self, self.__species__[x]+'_init') + '\n') if File != None: for x in out_list: File.write(x)
[docs] def showSpeciesFixed(self,File=None): """ showSpeciesFixed(File=None) Prints the current value of the model's fixed species values (mod.X) to screen or file. Arguments: File [default=None]: an open, writable Python file object """ out_list = [] print '\nFixed species' out_list.append('\n## fixed species\n') for x in range(len(self.__fixed_species__)): if File == None: print self.__fixed_species__[x] + ' = ' + self.__settings__['mode_number_format'] % getattr(self, self.__fixed_species__[x]) else: out_list.append(self.__fixed_species__[x] + ' = ' + self.__settings__['mode_number_format'] % getattr(self, self.__fixed_species__[x]) + '\n') if File != None: for x in out_list: File.write(x)
[docs] def showPar(self,File=None): """ showPar(File=None) Prints the current value of the model's parameter values (mod.P) to screen or file. Arguments: File [default=None]: an open, writable Python file object """ out_list = [] print '\nParameters' out_list.append('\n## parameters\n') for x in range(len(self.__parameters__)): if File == None: print self.__parameters__[x] + ' = ' + self.__settings__['mode_number_format'] % getattr(self, self.__parameters__[x]) else: out_list.append(self.__parameters__[x] + ' = ' + self.__settings__['mode_number_format'] % getattr(self, self.__parameters__[x]) + '\n') if File != None: for x in out_list: File.write(x)
[docs] def showModifiers(self,File=None): """ showModifiers(File=None) Prints the current value of the model's modifiers per reaction to screen or file. Arguments: File [default=None]: an open, writable Python file object """ noMod = [] out_list = [] print '\nModifiers per reaction:' out_list.append('\n## modifiers per reaction\n') for reac in self.__modifiers__: rstr = '' if len(reac[1]) == 1: if File == None: print reac[0] +' has modifier:', rstr = rstr + reac[0] +' has modifier: ' for x in reac[1]: if File == None: print x, rstr = rstr + x + ' ' if File == None: print ' ' rstr += '\n' elif len(reac[1]) > 1: if File == None: print reac[0] +' has modifiers: ', rstr = rstr + reac[0] +' has modifiers: ' for x in reac[1]: if File == None: print x, rstr = rstr + x + ' ' if File == None: print ' ' rstr += '\n' else: noMod.append(reac[0]) out_list.append(rstr) if len(noMod) > 0: print '\nReactions with no modifiers:' out_list.append('\n## reactions with no modifiers\n') cntr = 0 rstr = '' for n in range(len(noMod)): cntr += 1 if cntr > 6: if File == None: print ' ' rstr += '\n' cntr = 0 if File == None: print noMod[n], rstr = rstr + noMod[n] + ' ' if File == None: print ' ' rstr += '\n' out_list.append(rstr) if File != None: for x in out_list: File.write(x)
[docs] def showState(self,File=None): """ showState(File=None) Prints the result of the last steady-state analyses. Both steady-state flux's and species concentrations are shown. Arguments: File [default=None]: an open, writable Python file object """ out_list = [] print '\nSteady-state species concentrations' out_list.append('\n## Current steady-state species concentrations\n') if self.__StateOK__: for x in range(len(self.state_species)): if File == None: print self.__species__[x] + '_ss = ' + self.__settings__['mode_number_format'] % self.state_species[x] else: out_list.append(self.__species__[x] + '_ss = ' + self.__settings__['mode_number_format'] % self.state_species[x] + '\n') else: print 'No valid steady state found' out_list.append('No valid steady state found.\n') print '\nSteady-state fluxes' out_list.append('\n## Steady-state fluxes\n') if self.__StateOK__: for x in range(len(self.state_flux)): if File == None: print 'J_' + self.__reactions__[x] + ' = ' + self.__settings__['mode_number_format'] % self.state_flux[x] else: out_list.append('J_' + self.__reactions__[x] + ' = ' + self.__settings__['mode_number_format'] % self.state_flux[x] + '\n') else: print 'No valid steady state found' out_list.append('No valid steady state found.\n') if File != None: for x in out_list: File.write(x)
[docs] def showRate(self,File=None): """ Prints the current rates of all the reactions using the current parameter values and species concentrations - *File* an open, writable Python file object (default=None) """ Vtemp = [r() for r in getattr(self, self.__reactions__[r])] outrate = '' for x in range(len(self.__reactions__)): outrate += self.__reactions__[x] + ' = ' + self.__settings__['mode_number_format'] % Vtemp[x] + '\n' del s,Vtemp if File != None: print '\nReaction rates' File.write('\n##Reaction rates\n') File.write(outrate) else: print '\nReaction rates' print outrate
[docs] def showRateEq(self,File=None): """ showRateEq(File=None) Prints the reaction stoichiometry and rate equations to screen or File. Arguments: File [default=None]: an open, writable Python file object """ out_list = [] tnform = self.__settings__['mode_number_format'] self.__settings__['mode_number_format'] = '%2.5f' print '\nReaction stoichiometry and rate equations' out_list.append('\n## Reaction stoichiometry and rate equations\n') # writes these out in a better order for key in self.Kmatrix.row: if File == None: print key + ':' else: out_list.append(key + ':\n') reagL = [] reagR = [] for reagent in self.__nDict__[key]['Reagents']: if self.__nDict__[key]['Reagents'][reagent] > 0: if self.__nDict__[key]['Reagents'][reagent] == 1.0: reagR.append(reagent.replace('self.','')) else: reagR.append('{' + self.__settings__['mode_number_format'] % abs(self.__nDict__[key]['Reagents'][reagent]) + '}' + reagent.replace('self.','')) elif self.__nDict__[key]['Reagents'][reagent] < 0: if self.__nDict__[key]['Reagents'][reagent] == -1.0: reagL.append(reagent.replace('self.','')) else: reagL.append('{' + self.__settings__['mode_number_format'] % abs(self.__nDict__[key]['Reagents'][reagent]) + '}' + reagent.replace('self.','')) substring = '' count = 0 for x in reagL: if count != 0: substring += ' + ' substring += x.replace(' ','') count += 1 prodstring = '' count = 0 for x in reagR: if count != 0: prodstring += ' + ' prodstring += x.replace(' ','') count += 1 if self.__nDict__[key]['Type'] == 'Rever': symbol = ' = ' else: symbol = ' > ' if File == None: print '\t' + substring + symbol + prodstring print '\t' + self.__nDict__[key]['RateEq'].replace('self.','') else: out_list.append('\t' + substring + symbol + prodstring + '\n') out_list.append('\t' + self.__nDict__[key]['RateEq'].replace('self.','') + '\n\n') if len(self.__rules__.keys()) > 0: out_list.append('\n# Assignment rules\n') for ass in self.__rules__: out_list.append('!F %s = %s\n' % (self.__rules__[ass]['name'],\ self.__rules__[ass]['formula'])) out_list.append('\n') if File != None: for x in out_list: File.write(x) self.__settings__['mode_number_format'] = tnform
[docs] def showODE(self,File=None,fmt='%2.3f'): """ showODE(File=None,fmt='%2.3f') Print a representation of the full set of ODE's generated by PySCeS to screen or file. Arguments: File [default=None]: an open, writable Python file object fmt [default='%2.3f']: output number format """ maxmetlen = 0 for x in self.__species__: if len(x) > maxmetlen: maxmetlen = len(x) maxreaclen = 0 for x in self.__reactions__: if len(x) > maxreaclen: maxreaclen = len(x) odes = '\n## ODE\'s (unreduced)\n' for x in range(self.__nmatrix__.shape[0]): odes += 'd' + self.__species__[x] + (maxmetlen-len(self.__species__[x]))*' ' + '|' beginline = 0 for y in range(self.__nmatrix__.shape[1]): if abs(self.__nmatrix__[x,y]) > 0.0: if self.__nmatrix__[x,y] > 0.0: if beginline == 0: odes += ' ' + fmt % abs(self.__nmatrix__[x,y]) + '*' + self.__reactions__[y] + (maxreaclen-len(self.__reactions__[y]))*' ' beginline = 1 else: odes += ' + ' + fmt % abs(self.__nmatrix__[x,y]) + '*' + self.__reactions__[y] + (maxreaclen-len(self.__reactions__[y]))*' ' else: if beginline == 0: odes += ' -' + fmt % abs(self.__nmatrix__[x,y]) + '*' + self.__reactions__[y] + (maxreaclen-len(self.__reactions__[y]))*' ' else: odes += ' - ' + fmt % abs(self.__nmatrix__[x,y]) + '*' + self.__reactions__[y] + (maxreaclen-len(self.__reactions__[y]))*' ' beginline = 1 odes += '\n' if self.__HAS_RATE_RULES__: odes += "\n## Rate Rules:\n" for rule in self.__rate_rules__: odes += "d%s | %s\n" % (rule, self.__rules__[rule]['formula'].replace('()','')) odes += '\n' if File != None: print '\nODE\'s (unreduced)\n' File.write(odes) else: print odes
[docs] def showODEr(self,File=None,fmt='%2.3f'): """ showODEr(File=None,fmt='%2.3f') Print a representation of the reduced set of ODE's generated by PySCeS to screen or file. Arguments: File [default=None]: an open, writable Python file object fmt [default='%2.3f']: output number format """ maxmetlen = 0 for x in self.__species__: if len(x) > maxmetlen: maxmetlen = len(x) maxreaclen = 0 for x in self.__reactions__: if len(x) > maxreaclen: maxreaclen = len(x) odes = '\n## ODE\'s (reduced)\n' for x in range(self.nrmatrix.shape[0]): odes += 'd' + self.__species__[self.nrmatrix_row[x]] + (maxmetlen-len(self.__species__[self.nrmatrix_row[x]]))*' ' + '|' beginline = 0 for y in range(self.__nrmatrix__.shape[1]): if abs(self.__nrmatrix__[x,y]) > 0.0: if self.__nrmatrix__[x,y] > 0.0: if beginline == 0: odes += ' ' + fmt % abs(self.__nrmatrix__[x,y]) + '*' + self.__reactions__[self.nrmatrix_col[y]] + (maxreaclen-len(self.__reactions__[self.nrmatrix_col[y]]))*' ' beginline = 1 else: odes += ' + ' + fmt % abs(self.__nrmatrix__[x,y]) + '*' + self.__reactions__[self.nrmatrix_col[y]] + (maxreaclen-len(self.__reactions__[self.nrmatrix_col[y]]))*' ' else: if beginline == 0: odes += ' -' + fmt % abs(self.__nrmatrix__[x,y]) + '*' + self.__reactions__[self.nrmatrix_col[y]] + (maxreaclen-len(self.__reactions__[self.nrmatrix_col[y]]))*' ' else: odes += ' - ' + fmt % abs(self.__nrmatrix__[x,y]) + '*' + self.__reactions__[self.nrmatrix_col[y]] + (maxreaclen-len(self.__reactions__[self.nrmatrix_col[y]]))*' ' beginline = 1 odes += '\n' if self.__HAS_RATE_RULES__: odes += "\n## Rate Rules:\n" for rule in self.__rate_rules__: odes += "d%s | %s\n" % (rule, self.__rules__[rule]['formula'].replace('()','')) odes += '\n' if File != None: print '\nODE\'s (reduced)\n' File.write(odes) self.showConserved(File) else: print odes self.showConserved()
[docs] def showModel(self,filename=None,filepath=None,skipcheck=0): """ showModel(filename=None,filepath=None,skipcheck=0) The PySCeS 'save' command, prints the entire model to screen or File in a PSC format. (Currently this only applies to basic model attributes, ! functions are not saved). Arguments: filename [default=None]: the output PSC file filepath [default=None]: the output directory skipcheck [default=0]: skip check to see if the file exists (1) auto-averwrite """ if filepath == None: filepath = self.ModelDir if filename == None: print '\nFixed species' if len(self.__fixed_species__) == 0: print '<none>' else: for x in self.__fixed_species__: print x, print ' ' self.showRateEq() self.showSpeciesI() self.showPar() else: #assert type(filename) == str, 'showModel() takes a string filename argument' FBuf = 0 if type(filename) == str: #filename = str(filename) filename = chkpsc(filename) go = 0 loop = 0 if skipcheck != 0: loop = 1 go = 1 while loop == 0: try: filex = os.path.join(filepath,filename) f = open(filex,'r') f.close() input = raw_input('\nfile "' + filex + '" exists.\nOverwrite? ([y]/n) ') if input == 'y' or input == '': go = 1 loop = 1 elif input == 'n': filename = raw_input('\nfile "' + filename + '" exists. Enter a new filename: ') go = 1 filex = os.path.join(filepath,filename) filename = chkpsc(filename) else: print '\nInvalid input' except: print '\nfile "' + filex + '" does not exist, proceeding' loop = 1 go = 1 else: print '\nI hope we have a filebuffer' if type(filename)==file or type(filename)==cStringIO.OutputType: # --johann 20070615 go = 1 FBuf = 1 print 'Seems like it' else: go = 0 print 'Are you sure you know what ur doing' if go == 1: if not FBuf: if filex[-4:] == '.psc': pass else: print 'Assuming extension is .psc' filex += '.psc' outFile = open(filex,'w') else: outFile = filename #header = '# \n' header = '# PySCeS (' + __version__ + ') model input file \n' #header += '# Copyright Brett G. Olivier, 2004 (bgoli at sun dot ac dot za) \n' header += '# http://pysces.sourceforge.net/ \n' header += '# \n' header += '# Original input file: ' + self.ModelFile + '\n' header += '# This file generated: ' + time.strftime("%a, %d %b %Y %H:%M:%S") + '\n\n' outFile.write(header) outFile.write('## Fixed species\n') if len(self.__fixed_species__) == 0: outFile.write('# <none>') else: outFile.write('FIX: ') for x in self.__fixed_species__: outFile.write(x + ' ') outFile.write('\n') self.showRateEq(File=outFile) self.showSpeciesI(File=outFile) self.showPar(File=outFile) if type(filename)==str or type(filename)==file: # don't close cStringIO objects -- johann 20070615 outFile.close()
[docs] def showN(self,File=None,fmt='%2.3f'): """ showN(File=None,fmt='%2.3f') Print the stoichiometric matrix (N), including row and column labels to screen or File. Arguments: File [default=None]: an open, writable Python file object fmt [default='%2.3f']: output number format """ km_trunc = 0 km_trunc2 = 0 km = '' rtr = [] ctr = [] for a in range(len(self.nmatrix_col)): if a == 0: km += 9*' ' if len(self.__reactions__[self.nmatrix_col[a]]) > 7: km += self.__reactions__[self.nmatrix_col[a]][:6]+'* ' km_trunc = 1 ctr.append(a) else: km += self.__reactions__[a] + (9-len(self.__reactions__[a])-1)*' ' for x in range(len(self.nmatrix_row)): if len(self.__species__[x]) > 6: km += '\n' + self.__species__[x][:5]+'*'+1*' ' km_trunc2 = 1 rtr.append(x) else: km += '\n' + self.__species__[x] + (8-len(self.__species__[x])-1)*' ' #km += '\n' + self.__reactions__[self.nmatrix_row[x]] + 4*' ' for y in range(len(self.nmatrix_col)): if y != 0: km += 3*' ' if self.__nmatrix__[x,y] >= 10.0: km += ' ' + fmt % self.__nmatrix__[x,y] elif self.__nmatrix__[x,y] >= 0.0: km += ' ' + fmt % self.__nmatrix__[x,y] else: if abs(self.__nmatrix__[x,y]) >= 10.0: km += fmt % self.__nmatrix__[x,y] else: km += ' ' + fmt % self.__nmatrix__[x,y] if km_trunc: km += '\n\n' + '* indicates truncated (col) reaction\n(' for x in range(len(self.nmatrix_col)): try: ctr.index(x) km += self.__reactions__[x] + ', ' except: pass km += ')' if km_trunc2: km += '\n\n' + '* indicates truncated (row) species:\n(' for x in range(len(self.nmatrix_row)): try: rtr.index(x) km += self.__species__[x] + ', ' except: pass km += ')' if File != None: print '\nStoichiometric matrix (N)' File.write('\n\n## Stoichiometric matrix (N)\n') File.write(km) File.write('\n') else: print '\nStoichiometric matrix (N)' print km
[docs] def showNr(self,File=None,fmt='%2.3f'): """ showNr(File=None,fmt='%2.3f') Print the reduced stoichiometric matrix (Nr), including row and column labels to screen or File. Arguments: File [default=None]: an open, writable Python file object fmt [default='%2.3f']: output number format """ km_trunc = 0 km_trunc2 = 0 km = '' ctr = [] rtr = [] for a in range(len(self.nrmatrix_col)): if a == 0: km += 9*' ' if len(self.__reactions__[self.nrmatrix_col[a]]) > 7: km += self.__reactions__[self.nrmatrix_col[a]][:6]+'* ' km_trunc = 1 ctr.append(a) else: km += self.__reactions__[a] + (9-len(self.__reactions__[a])-1)*' ' for x in range(len(self.nrmatrix_row)): if len(self.__species__[self.nrmatrix_row[x]]) > 6: km += '\n' + self.__species__[self.nrmatrix_row[x]][:5]+'*'+1*' ' km_trunc2 = 1 rtr.append(x) else: km += '\n' + self.__species__[self.nrmatrix_row[x]] + (8-len(self.__species__[self.nrmatrix_row[x]])-1)*' ' #km += '\n' + self.__reactions__[self.nrmatrix_row[x]] + 4*' ' for y in range(len(self.nrmatrix_col)): if y != 0: km += 3*' ' if self.__nrmatrix__[x,y] >= 10.0: km += ' ' + fmt % self.__nrmatrix__[x,y] elif self.__nrmatrix__[x,y] >= 0.0: km += ' ' + fmt % self.__nrmatrix__[x,y] else: if abs(self.__nrmatrix__[x,y]) >= 10.0: km += fmt % self.__nrmatrix__[x,y] else: km += ' ' + fmt % self.__nrmatrix__[x,y] if km_trunc: km += '\n\n' + '* indicates truncated (col) reaction:\n(' for x in range(len(self.nrmatrix_col)): try: ctr.index(x) km += self.__reactions__[x] + ', ' except: pass km += ')' if km_trunc2: km += '\n\n' + '* indicates truncated (row) species:\n(' for x in range(len(self.nrmatrix_row)): try: rtr.index(x) km += self.__species__[self.nrmatrix_row[x]] + ', ' except: pass km += ')' if File != None: print '\nReduced stoichiometric matrix (Nr)' File.write('\n\n## Reduced stoichiometric matrix (Nr)\n') File.write(km) File.write('\n') else: print '\nReduced stoichiometric matrix (Nr)' print km
[docs] def showK(self,File=None,fmt='%2.3f'): """ showK(File=None,fmt='%2.3f') Print the Kernel matrix (K), including row and column labels to screen or File. Arguments: File [default=None]: an open, writable Python file object fmt [default='%2.3f']: output number format """ km_trunc = 0 km_trunc2 = 0 km = '' ctr = [] rtr = [] for a in range(len(self.kmatrix_col)): if a == 0: km += 8*' ' if len(self.__reactions__[self.kmatrix_col[a]]) > 7: km += self.__reactions__[self.kmatrix_col[a]][:6]+'* ' km_trunc = 1 ctr.append(a) else: km += self.__reactions__[self.kmatrix_col[a]] + (9-len(self.__reactions__[self.kmatrix_col[a]])-1)*' ' for x in range(len(self.kmatrix_row)): if len(self.__reactions__[self.kmatrix_row[x]]) > 6: km += '\n' + self.__reactions__[self.kmatrix_row[x]][:5]+'*'+1*' ' km_trunc2 = 1 rtr.append(x) else: km += '\n' + self.__reactions__[self.kmatrix_row[x]] + (8-len(self.__reactions__[self.kmatrix_row[x]])-1)*' ' #km += '\n' + self.__reactions__[self.kmatrix_row[x]] + 4*' ' for y in range(len(self.kmatrix_col)): if y != 0: km += 4*' ' if abs(self.__kmatrix__[x,y]) == 0.0: km += ' ' + fmt % abs(self.__kmatrix__[x,y]) elif self.__kmatrix__[x,y] < 0.0: km += fmt % self.__kmatrix__[x,y] else: km += ' ' + fmt % self.__kmatrix__[x,y] if km_trunc: km += '\n\n' + '* indicates truncated (col) reaction:\n(' for x in range(len(self.kmatrix_col)): try: ctr.index(x) km += self.__reactions__[self.kmatrix_col[x]] + ', ' except: pass km += ')' if km_trunc2: km += '\n\n' + '* indicates truncated (row) reaction:\n(' for x in range(len(self.kmatrix_row)): try: rtr.index(x) km += self.__reactions__[self.kmatrix_row[x]] + ', ' except: pass km += ')' if File != None: print '\nKernel matrix (K)' File.write('\n\n## Kernel matrix (K)\n') if not self.__HAS_FLUX_CONSERVATION__: File.write('No flux conservation\n') else: File.write(km) File.write('\n') else: print '\nKernel matrix (K)' if not self.__HAS_FLUX_CONSERVATION__: print 'No flux conservation' else: print km
[docs] def showL(self,File=None,fmt='%2.3f'): """ showL(File=None,fmt='%2.3f') Print the Link matrix (L), including row and column labels to screen or File. Arguments: File [default=None]: an open, writable Python file object fmt [default='%2.3f']: output number format """ km_trunc = 0 km_trunc2 = 0 km = '' ctr = [] rtr = [] for a in range(len(self.lmatrix_col)): if a == 0: km += 8*' ' if len(self.__species__[self.lmatrix_col[a]]) > 7: km += self.__species__[self.lmatrix_col[a]][:6]+'* ' km_trunc = 1 ctr.append(a) else: km += self.__species__[self.lmatrix_col[a]] + (9-len(self.__species__[self.lmatrix_col[a]])-1)*' ' for x in range(len(self.lmatrix_row)): if len(self.__species__[self.lmatrix_row[x]]) > 6: km += '\n' + self.__species__[self.lmatrix_row[x]][:5]+'* ' km_trunc2 = 1 rtr.append(x) else: km += '\n' + self.__species__[self.lmatrix_row[x]] + (8-len(self.__species__[self.lmatrix_row[x]])-1)*' ' for y in range(len(self.lmatrix_col)): if y != 0: km += 4*' ' if abs(self.__lmatrix__[x,y]) == 0.0: km += ' ' + fmt % abs(self.__lmatrix__[x,y]) elif self.__lmatrix__[x,y] < 0.0: km += fmt % self.__lmatrix__[x,y] else: km += ' ' + fmt % self.__lmatrix__[x,y] if km_trunc: km += '\n\n' + '* indicates truncated (col) species:\n(' for x in range(len(self.lmatrix_col)): try: ctr.index(x) km += self.__species__[self.lmatrix_col[x]] + ', ' except: pass km += ')' if km_trunc2: km += '\n\n' + '* indicates truncated (row) species:\n(' for x in range(len(self.lmatrix_row)): try: rtr.index(x) km += self.__species__[self.lmatrix_row[x]] + ', ' except: pass km += ')' if File != None: print '\nLink matrix (L)' File.write('\n\n## Link matrix (L)\n') if not self.__HAS_MOIETY_CONSERVATION__: File.write('No moiety conservation\n') File.write(km) File.write('\n') else: File.write(km) File.write('\n') else: print '\nLink matrix (L)' if not self.__HAS_MOIETY_CONSERVATION__: print '\"No moiety conservation\"\n' print km else: print km #Internal test functions
[docs] def TestSimState(self,endTime=10000,points=101,diff=1.0e-5): """ **Deprecated** """ pass
[docs] def ResetNumberFormat(self): """ ResetNumberFormat() Reset PySCeS default number format stored as mod.mode_number format to %2.4e Arguments: None """ self.__settings__['mode_number_format'] = '%2.4e'
[docs] def Write_array(self,input,File=None,Row=None,Col=None,close_file=0,separator=' '): """ Write_array(input,File=None,Row=None,Col=None,close_file=0,separator=' ') Write an array to File with optional row/col labels. A ',' separator can be specified to create a CSV style file. mod.__settings__['write_array_header']: add <filename> as a header line (1 = yes, 0 = no) mod.__settings__['write_array_spacer']: add a space after the header line (1 = yes, 0 = no) mod.__settings__['write_arr_lflush']: set the flush rate for large file writes Arguments: input: the array to be written File [default=None]: an open, writable Python file object Row [default=None]: a list of row labels Col [default=None]: a list of column labels close_file [default=0]: close the file after write (1) or leave open (0) separator [default=' ']: the column separator to use """ fname = 'Write_array_' + self.ModelFile.replace('.psc','') + '_' + time.strftime("%H:%M:%S") # seeing as row indexes are now tuples and not arrays - brett 20050713 if type(input) == tuple or type(input) == list: input = numpy.array(input) if Row != None: assert len(Row) == input.shape[0], 'len(Row) must be equal to len(input_row)' if Col != None: assert len(Col) == input.shape[1], 'len(Col) must be equal to len(input_col)' if File != None: #assert File != file, 'WriteArray(input,File=None,Row=None,Col=None,close_file=0)' if self.__settings__['write_array_header']: File.write('\n## ' + fname + '\n') if self.__settings__['write_array_spacer']: File.write('\n') if Col != None: print 'Writing column' File.write('#') try: input_width = len(self.__settings__['mode_number_format'] % input[0,0])+ 1 + len(str(separator)) except: input_width = len(self.__settings__['mode_number_format'] % input[0])+ 1 + len(str(separator)) for x in Col: if len(str(x)) <= len(str(input[0,0])): spacer = (input_width-len(str(x)))*' ' File.write(str(x) + spacer) else: spacer = (len(str(separator)))*' ' File.write(str(x[:input_width]) + spacer) File.write('\n') try: print '\nWriting array (normal) to file' #scipy.io.write_array(File,input,separator=' ',linesep='\n',precision=3,keep_open=1) flush_count = 0 if self.__settings__['write_arr_lflush'] < 2: print 'INFO: LineFlush must be >= 2' self.__settings__['write_arr_lflush'] = 2 for x in range(input.shape[0]): flush_count += 1 for y in range(input.shape[1]): if input[x,y] < 0.0: File.write(self.__settings__['mode_number_format'] % input[x,y]) else: File.write(' ' + self.__settings__['mode_number_format'] % input[x,y]) if y < input.shape[1]-1: File.write(separator) File.write('\n') if flush_count == self.__settings__['write_arr_lflush']: File.flush() flush_count = 0 #print 'flushing' #File.write('\n') except IndexError, e: print '\nWriting vector (normal) to file' for x in range(len(input)): File.write(self.__settings__['mode_number_format'] % input[x]) if x < len(input)-1: File.write(separator) File.write('\n') if Row != None: print 'Writing row' File.write('# Row: ') for x in Row: File.write(str(x)+separator) File.write('\n') if close_file: File.close() else: print 'INFO: You need to supply an open writable file as the 2nd argument to this function'
[docs] def Write_array_latex(self,input,File=None,Row=None,Col=None,close_file=0): """ Write_array_latex(input,File=None,Row=None,Col=None,close_file=0) Write an array to an open file as a \'LaTeX\' {array} Arguments: input: the array to be written File [default=None]: an open, writable Python file object Row [default=None]: a list of row labels Col [default=None]: a list of column labels close_file [default=0]: close the file after write (1) or leave open (0) """ # seeing as row indexes are now tuples and not arrays - brett 20050713 if type(input) == tuple or type(input) == list: input = numpy.array(input) try: a = input.shape[1] except: input.shape = (1,input.shape[0]) if Row != None: assert len(Row) == input.shape[0], 'len(Row) must be equal to len(input_row)' print 'Writing row' if Col != None: assert len(Col) == input.shape[1], 'len(Col) must be equal to len(input_col)' print 'Writing column' if self.__settings__['write_arr_lflush'] < 2: print 'INFO: LineFlush must be >= 2' self.__settings__['write_arr_lflush'] = 2 fname = 'Write_array_latex_' + self.ModelFile.replace('.psc','') + '_' + time.strftime("%H:%M:%S") if File != None: #assert File != file, 'WriteArray(input,File=None,Row=None,Col=None,close_file=0)' File.write('\n%% ' + fname + '\n') try: a = input.shape print '\nWriting array (LaTeX) to file' File.write('\\[\n') File.write('\\begin{array}{') if Row != None: File.write('r|') for x in range(input.shape[1]): File.write('r') File.write('}\n ') flush_count = 0 for x in range(input.shape[0]): if Col != None and x == 0: for el in range(len(Col)): elx = str(Col[el]).replace('_','\_') if Row != None: File.write(' & $\\small{' + elx + '}$') else: if el == 0: File.write(' $\\small{' + elx + '}$') else: File.write(' & $\\small{' + elx + '}$') File.write(' \\\\ \\hline\n ') flush_count += 1 if Row != None: el2 = str(Row[x]).replace('_','\\_') File.write('$\\small{' + el2 + '}$ &') for y in range(input.shape[1]): if input[x,y] < 0.0: val = '%2.4f' % input[x,y] else: val = ' ' + '%2.4f' % input[x,y] File.write(val) if y < input.shape[1]-1: File.write(' &') File.write(' \\\\\n ') if flush_count == self.__settings__['write_arr_lflush']: File.flush() flush_count = 0 #print 'flushing' #File.write('\n') File.write('\\end{array}\n') File.write('\\]\n\n') except: print '\nINFO: Only arrays can currently be processed with this method.\n' if close_file: File.close() else: print 'INFO: You need to supply an open writable file as the 2nd argument to this method'
[docs] def Write_array_html(self,input,File=None,Row=None,Col=None,name=None,close_file=0): """ Write_array_html(input,File=None,Row=None,Col=None,name=None,close_file=0) Write an array as an HTML table (no header/footer) or complete document. Tables are formatted with coloured columns if they exceed a specified size. mod.__settings__['write_array_html_header']: write the HTML document header mod.__settings__['write_array_html_footer']: write the HTML document footer Arguments: input: the array to be written File [default=None]: an open, writable Python file object Row [default=None]: a list of row labels Col [default=None]: a list of column labels name [default=None]: an HTML table description line close_file [default=0]: close the file after write (1) or leave open (0) """ # seeing as row indexes are now tuples and not arrays - brett 20050713 if type(input) == tuple or type(input) == list: input = numpy.array(input) try: a = input.shape[1] except: input.shape = (1,input.shape[0]) if Row != None: assert len(Row) == input.shape[0], 'len(Row) must be equal to len(input_row)' print 'Writing row' if Col != None: assert len(Col) == input.shape[1], 'len(Col) must be equal to len(input_col)' print 'Writing column' if self.__settings__['write_arr_lflush'] < 2: print 'INFO: LineFlush must be >= 2' self.__settings__['write_arr_lflush'] = 2 if name != None: fname = 'PySCeS data "'+name+'" generated from model file: '+self.ModelFile+ ' ' + time.strftime("%H:%M:%S") else: fname = 'PySCeS data generated from model file: '+self.ModelFile+' '+time.strftime("%H:%M:%S") if File != None: #assert File != file, 'WriteArray(input,File=None,Row=None,Col=None,close_file=0)' header = '\n' header += '<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">\n' header += '<html>\n' header += '<head>\n' header += '<title>PySCeS data generated from: ' + self.ModelFile + ' - ' + time.strftime("%H:%M:%S (%Z)") + '</title>\n' header += '<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">\n' header += '</head>\n' header += '<body>\n\n' header += '<h4><a href="http://pysces.sourceforge.net">PySCeS</a> data generated from: ' + self.ModelFile + '</h4>\n\n' if self.__settings__['write_array_html_header']: File.write(header) File.write('<!-- ' + fname + '-->\n') if name != None: File.write('\n<p>\n<font face="Arial, Helvetica, sans-serif"><strong>' + name + '</strong></font>') else: File.write('\n<p>\n') File.write('\n<table border="1" cellpadding="2" cellspacing="2" bgcolor="#FFFFFF">') try: a = input.shape print 'Writing array (HTML) to file' double_index = 15 if Col != None: File.write('\n<tr>\n') if Row != None: File.write(' <td>&nbsp;</td>\n') for col in Col: File.write(' <td bgcolor="#CCCCCC"><div align="center"><b>' + str(col) + '</b></div></td>\n') if Row != None and input.shape[1]+1 >= double_index: File.write(' <td>&nbsp;</td>\n') File.write('</tr>') flush_count = 0 colour_count_row = 6 colour_count_col = 6 rowcntr = 0 html_colour = '#FFFFCC' for x in range(input.shape[0]): rowcntr += 1 if rowcntr == colour_count_row and input.shape[0] > 3*colour_count_row: File.write('\n<tr bgcolor="'+html_colour+'">\n') rowcntr = 0 else: File.write('\n<tr>\n') if Row != None: File.write(' <td bgcolor="#CCCCCC"><div align="center"><b>' + str(Row[x]) + '</b></div></td>\n') colcntr = 0 for y in range(input.shape[1]): colcntr += 1 if colcntr == colour_count_col and input.shape[1] > 3*colour_count_row: File.write(' <td nowrap bgcolor="'+html_colour+'">' + self.__settings__['write_array_html_format'] % input[x,y] + '</td>\n') colcntr = 0 else: File.write(' <td nowrap>' + self.__settings__['write_array_html_format'] % input[x,y] + '</td>\n') ## File.write(' <td nowrap>' + self.__settings__['write_array_html_format'] % input[x,y] + '</td>\n') if Row != None and input.shape[1]+1 >= double_index: File.write(' <td bgcolor="#CCCCCC"><div align="center"><b>' + Row[x] + '</b></div></td>\n') File.write('</tr>') if flush_count == self.__settings__['write_arr_lflush']: File.flush() flush_count = 0 #print 'flushing' if Col != None and input.shape[0]+1 >= double_index: File.write('\n<tr>\n') if Row != None: File.write(' <td>&nbsp;</td>\n') for col in Col: File.write(' <td bgcolor="#CCCCCC"><div align="center"><b>' + col + '</b></div></td>\n') if Row != None and input.shape[1]+1 >= double_index: File.write(' <td>&nbsp;</td>\n') File.write('</tr>\n') except: print '\nINFO: Only arrays can currently be processed with this method.\n' File.write('\n</table>\n') File.write('</p>\n\n') if self.__settings__['write_array_html_footer']: try: File.write('<p><a href="http://pysces.sourceforge.net"><font size="3">PySCeS '+__version__+\ '</font></a><font size="2"> HTML output (model <i>'+self.ModelFile+\ '</i> analysed at '+time.strftime("%H:%M:%S")+' by <i>'+getuser()+'</i>)</font></p>\n') except: File.write('<p><a href="http://pysces.sourceforge.net"><font size="3">PySCeS '+__version__+\ '</font></a><font size="2"> HTML output (model <i>'+self.ModelFile+\ '</i> analysed at '+time.strftime("%H:%M:%S - %Z")+')</font></p>\n') File.write('</body>\n') File.write('</html>\n') else: File.write('\n') if close_file: File.close() else: print 'INFO: You need to supply an open writable file as the 2nd argument to this method'
[docs] def SimPlot(self, plot='species', filename=None, title=None, log=None, format='lines'): """ Plot the simulation results, uses the new UPI pysces.plt interface: - *plot*: output to plot (default='species') - 'all' rates and species - 'species' species - 'rates' reaction rates - `['S1', 'R1', ]` a list of model attributes (species, rates) - *filename* if not None file is exported to filename (default=None) - *title* the plot title (default=None) - *log* use log axis for 'x', 'y', 'xy' (default=None) - *format* line format, backend dependant (default='') """ data = None labels = None allowedplots = ['all', 'species', 'rates'] if type(plot) != list and plot not in allowedplots: raise RuntimeError, '\nPlot must be one of %s not \"%s\"' % (str(allowedplots), plot) if plot == 'all': data, labels = self.data_sim.getAllSimData(lbls=True) elif plot == 'species': data, labels = self.data_sim.getSpecies(lbls=True) elif plot == 'rates': data, labels = self.data_sim.getRates(lbls=True) else: plot = [at for at in plot if at in self.__species__+self.__reactions__+[self.data_sim.time_label]] kwargs = {'lbls' : True} if len(plot) > 0: data, labels = self.data_sim.getSimData(*plot, **kwargs) del allowedplots self.__SIMPLOT_OUT__ = labels plt.plotLines(data, 0, range(1, data.shape[1]), titles=labels, formats=[format]) # set the x-axis range so that it is original range + 0.2*sim_end # this is a sceintifcally dtermned amount of space that is needed for the title at the # end of the line :-) - brett 20040209 RngTime = self.data_sim.getTime() end = RngTime[-1] + 0.2*RngTime[-1] plt.setRange('x', RngTime[0], end) del RngTime if self.__KeyWords__['Output_In_Conc']: M = 'Concentration' else: M = 'Amount (%(multiplier)s x %(kind)s x 10**%(scale)s)**%(exponent)s' % self.__uDict__['substance'] xu = 'Time (%(multiplier)s x %(kind)s x 10**%(scale)s)**%(exponent)s' % self.__uDict__['time'] plt.setAxisLabel('x', xu) if plot == 'all': yl = 'Rates, %s' % M elif plot == 'rates': yl = 'Rate' elif plot == 'species': yl = '%s' % M else: yl = 'Rates, %s' % M plt.setAxisLabel('y', yl) if log != None: plt.setLogScale(log) if title == None: plt.setGraphTitle('PySCeS Simulation (' + self.ModelFile + ') ' + time.strftime("%a, %d %b %Y %H:%M:%S")) else: plt.setGraphTitle(title) plt.replot() if filename != None: plt.export(filename, directory=self.ModelOutput, type='png')
[docs] def Scan1(self,range1=[],runUF=0): """ Scan1(range1=[],runUF=0) Perform a single dimension parameter scan using the steady-state solvers. The parameter to be scanned is defined (as a model attribute "P") in mod.scan_in while the required output is entered into the list mod.scan_out. Results of a parameter scan can be easilly viewed with Scan1Plot(). mod.scan_in - a model attribute written as in the input file (eg. P, Vmax1 etc) mod.scan_out - a list of required output ['A','T2', 'ecR1_s1', 'ccJR1_R1', 'rcJR1_s1', ...] mod.scan_res - the results of a parameter scan mod.scan - numpy record array with the scan results (scan_in and scan_out), call as mod.scan.Vmax, mod.scan.A_ss, mod.scan.J_R1, etc. mod.__settings__["scan1_mca_mode"] - force the scan algorithm to evaluate the elasticities (1) and control coefficients (2) (this should also be auto-detected by the Scan1 method). Arguments: range1 [default=[]]: a predefined range over which to scan. runUF [default=0]: run (1) the user defined function mod.User_Function (!U) before evaluating the steady state. """ if self.__settings__["scan1_dropbad"] != 0: print '\n****\nINFO: Dropping invalid steady states can mask interesting behaviour\n****\n' #check the legitimacy of the input and output lists #self.__settings__["scan1_mca_mode"] = scanMCA run = 1 # we now accept parameters and intial species values as scanable values - brett 20060519 parameters = list(self.__parameters__) + [a+'_init' for a in self.__species__] scanpar = [] scanIn = [self.scan_in] for x in scanIn: try: a = parameters.index(x) scanpar.append(x) except: print x, ' is not a parameter' if len(scanpar) < 1: print 'mod.scan_in should contain something' run = 0 elif len(scanpar) > 1: print 'Only 1D scans possible - first element will be used' scanpar = scanpar[0] print'self.scan_in = ', scanpar else: scanpar = scanpar[0] self.scan_in = scanpar wrong = [] for x in self.scan_out: if x[:2] == 'ec' and self.__settings__["scan1_mca_mode"] < 2: self.__settings__["scan1_mca_mode"] = 1 elif x[:2] == 'cc': self.__settings__["scan1_mca_mode"] = 2 elif x[:2] == 'rc': self.__settings__["scan1_mca_mode"] = 3 if self.__settings__["scan1_mca_mode"] == 1: self.doElas() elif self.__settings__["scan1_mca_mode"] == 2: self.doMca() elif self.__settings__["scan1_mca_mode"] == 3: self.doMcaRC() else: self.doState() # we are going to run a whole bunch anyway for x in self.scan_out: try: if x.startswith('ec'): try: getattr(self.ec, x[2:]) except AttributeError: getattr(self.ecp, x[2:]) elif x.startswith('cc'): if x.startswith('ccJ'): getattr(self.cc, x[3:]) else: getattr(self.cc, x[2:]) elif x.startswith('rc'): if x.startswith('rcJ'): getattr(self.rc, x[3:]) else: getattr(self.rc, x[2:]) elif x in self.reactions: getattr(self.data_sstate, x) print "INFO: using steady-state flux for reaction (%s --> J_%s)" % (x, x) self.scan_out[self.scan_out.index(x)] = 'J_%s' % x elif x.startswith('J_'): getattr(self.data_sstate, x[2:]) elif x in self.species: getattr(self.data_sstate, x) print "INFO: using steady-state concentration for species (%s --> %s_ss)" % (x, x) self.scan_out[self.scan_out.index(x)] = '%s_ss' % x elif x.endswith('_ss'): getattr(self.data_sstate, x[:-3]) else: getattr(self, x) except: print x + ' is not a valid attribute' wrong.append(x) if x == self.scan_in: wrong.append(x) print x, ' is mod.scan_in ' if len(wrong) != 0: try: for x in wrong: print self.scan_out self.scan_out.remove(x) print self.scan_out except: print 'No valid output' run = 0 assert len(self.scan_out) != 0, "Output parameter list (mod.scan_out) empty - do the model attributes exist ... see manual for details." #print run, self.scan_in, self.scan_out self._scan = None result = [] cntr = 0 cntr2 = 1 cntr3 = 0 # reset scan error controls self.__scan_errors_par__ = None self.__scan_errors_idx__ = None if self.__settings__['scan1_mesg']: print '\nScanning ...' if len(self.scan_in) > 0 and run == 1: badList = [] badList_idx = [] if self.__settings__['scan1_mesg']: print len(range1)-(cntr*cntr2), for xi in range(len(range1)): x = range1[xi] setattr(self, self.scan_in, x) ## exec('self.' + self.scan_in + ' = ' + `x`) if self.__settings__["scan1_mca_mode"] == 1: self.doElas() elif self.__settings__["scan1_mca_mode"] == 2: self.doMca() elif self.__settings__["scan1_mca_mode"] == 3: self.doMcaRC() else: self.State() rawres = [x] # these two lines are going to be terminated - brett #rawres.append(x) #rawres.insert(0,x) if runUF: self.User_Function() for res in self.scan_out: if res.startswith('ec'): try: rawres.append(getattr(self.ec, res[2:])) except AttributeError: rawres.append(getattr(self.ecp, res[2:])) elif res.startswith('cc'): if res.startswith('ccJ'): rawres.append(getattr(self.cc, res[3:])) else: rawres.append(getattr(self.cc, res[2:])) elif res.startswith('rc'): if res.startswith('rcJ'): rawres.append(getattr(self.rc, res[3:])) else: rawres.append(getattr(self.rc, res[2:])) elif res in self.reactions: rawres.append(getattr(self.data_sstate, res)) elif res.startswith('J_'): rawres.append(getattr(self.data_sstate, res[2:])) elif res in self.species: rawres.append(getattr(self.data_sstate, res)) elif res.endswith('_ss'): rawres.append(getattr(self.data_sstate, res[:-3])) else: rawres.append(getattr(self, res)) # The following is for user friendly reporting: # next we check if the state is ok : if bad report it and add it : if good add it if not self.__StateOK__ and self.__settings__["scan1_dropbad"] != 0: pass elif not self.__StateOK__: if self.__settings__["scan1_nan_on_bad"]: result.append([numpy.NaN]*len(rawres)) else: result.append(rawres) badList.append(x) badList_idx.append(xi) else: result.append(rawres) cntr += 1 cntr3 += 1 if cntr == 20: if self.__settings__['scan1_mesg']: print len(range1)-(cntr*cntr2), cntr = 0 cntr2 += 1 if cntr3 == 101: if self.__settings__['scan1_mesg']: print ' ' cntr3 = 0 if self.__settings__['scan1_mesg']: print '\ndone.\n' if len(badList) != 0: self.__scan_errors_par__ = badList self.__scan_errors_idx__ = badList_idx print '\nINFO: ' + str(len(badList)) + ' invalid steady states detected at ' + self.scan_in + ' values:' print badList if len(result) == 0: self.scan_res = numpy.zeros((len(range1),len(self.scan_out)+1)) else: self.scan_res = numpy.array(result)
@property def scan(self): if self._scan is None and self.scan_res is not None: self._scan = numpy.rec.fromrecords(self.scan_res, names=[self.scan_in]+self.scan_out) return self._scan
[docs] def Scan1Plot(self, plot=[], title=None, log=None, format='lines', filename=None): """ Plot the results of a parameter scan generated with **Scan1()** - *plot* if empty mod.scan_out is used, otherwise any subset of mod.scan_out (default=[]) - *filename* the filename of the PNG file (default=None, no export) - *title* the plot title (default=None) - *log* if None a linear axis is assumed otherwise one of ['x','xy','xyz'] (default=None) - *format* the backend dependent line format (default='lines') or the *CommonStyle* 'lines' or 'points'. """ data = self.scan_res labels = None if type(plot) == str: plot = [plot] if len(plot) == 0: plot = self.scan_out else: plot = [at for at in plot if hasattr(self, at)] if len(plot) == 0: print 'No plottable output specified using self.scan_out' plot = self.scan_out plotidx = [self.scan_out.index(c)+1 for c in plot] labels = [self.scan_in] + self.scan_out plt.plotLines(data, 0, plotidx, titles=labels, formats=[format]) end = data[-1,0] + 0.2*data[-1,0] plt.setRange('x', data[0,0], end) plt.setAxisLabel('x', self.scan_in) yl = 'Steady-state variable' plt.setAxisLabel('y', yl) if log != None: plt.setLogScale(log) if title == None: plt.setGraphTitle('PySCeS Scan1 (' + self.ModelFile + ') ' + time.strftime("%a, %d %b %Y %H:%M:%S")) else: plt.setGraphTitle(title) plt.replot() if filename != None: plt.export(filename, directory=self.ModelOutput, type='png')
[docs] def Scan2D(self, p1, p2, output, log=False): """ Generate a 2 dimensional parameter scan using the steady-state solvers. - *p1* is a list of [parameter1, start, end, points] - *p2* is a list of [parameter2, start, end, points] - *output* steady-state variable/properties e.g. 'J_R1', 'A_ss', 'ecR1_s1' - *log* scan using log ranges for both axes """ for p in [p1, p2]: if not hasattr(self, p[0]): raise RuntimeError, '\"%s\" is not a valid model attribute' % p[0] p1 = list(p1) + [log, False] p2 = list(p2) + [log, False] sc1 = Scanner(self) sc1.quietRun=True sc1.addScanParameter(*p1) sc1.addScanParameter(*p2) sc1.addUserOutput(output) sc1.Run() self.__scan2d_pars__ = [p1[0], p2[0], output] self.__scan2d_results__ = sc1.getResultMatrix() del sc1
[docs] def Scan2DPlot(self, title=None, log=None, format='lines', filename=None): """ Plot the results of a 2D scan generated with Scan2D - *filename* the filename of the PNG file (default=None, no export) - *title* the plot title (default=None) - *log* if None a linear axis is assumed otherwise one of ['x','xy','xyz'] (default=None) - *format* the backend dependent line format (default='lines') or the *CommonStyle* 'lines' or 'points'. """ plt.splot(self.__scan2d_results__, 0, 1, 2, titles=self.__scan2d_pars__, format=format) plt.setRange('xyz') plt.setAxisLabel('x', self.__scan2d_pars__[0]) plt.setAxisLabel('y', self.__scan2d_pars__[1]) plt.setAxisLabel('z', 'Steady-state variable') if log != None: plt.setLogScale(log) if title == None: plt.setGraphTitle('PySCeS Scan2D (' + self.ModelFile + ') ' + time.strftime("%a, %d %b %Y %H:%M:%S")) else: plt.setGraphTitle(title) plt.replot() if filename != None: plt.export(filename, directory=self.ModelOutput, type='png')
[docs] def SetQuiet(self): """ SetQuiet() Turn off as much solver reporting noise as possible: mod.__settings__['hybrd_mesg'] = 0 mod.__settings__['nleq2_mesg'] = 0 mod.__settings__["lsoda_mesg"] = 0 mod.__settings__['mode_state_mesg'] = 0 mod.__settings__['solver_switch_warning'] = False Arguments: None """ self.__settings__['hybrd_mesg'] = 0 self.__settings__['nleq2_mesg'] = 0 self.__settings__["lsoda_mesg"] = 0 self.__settings__['mode_state_mesg'] = 0 self.__settings__['solver_switch_warning'] = False
[docs] def SetLoud(self): """ SetLoud() Turn on as much solver reporting noise as possible: mod.__settings__['hybrd_mesg'] = 1 mod.__settings__['nleq2_mesg'] = 1 mod.__settings__["lsoda_mesg"] = 1 mod.__settings__['mode_state_mesg'] = 1 mod.__settings__['solver_switch_warning'] = True Arguments: None """ self.__settings__['hybrd_mesg'] = 1 self.__settings__['nleq2_mesg'] = 1 self.__settings__["lsoda_mesg"] = 1 self.__settings__['mode_state_mesg'] = 1 self.__settings__['solver_switch_warning'] = True
[docs] def clone(self): """ Returns a deep copy of this model object (experimental!) """ print '\nINFO: Cloning function is currently experimental ... use with care.' return copy.deepcopy(self)
[docs]class WasteManagement(object): _gc_delay_hack = gplt # dont f$%^&*)ing even ask ... only 5 hrs to hack this workaround
__waste_manager = WasteManagement() if __psyco_active__: psyco.bind(PysMod) if __name__ == '__main__': print '\nTo use PySCeS import it from a Python Shell: \n\timport pysces\n' raw_input('Press <enter> to continue')