PyErf¶
A Pure-Python Error Function and Inverse Error Function Package¶
pyerf is a pure-Python implementation of the error function and
inverse error function using the same functions that SciPy uses (namely
parts of the Cephes math library, cprob/ndtr.c and cprob/ndtri.c).
This is a useful package for when you need to calculate some error fuctions but you don’t want to install all of the SciPy/NumPy stuff.
Usage¶
You can import the module:
from pyerf import pyerf
pyerf.erfinv(0.5) # 0.476936...
pyerf.erf(0.5) # 0.5204998...
pyerf.erfc(0.5) # 0.4795001...
or the package:
import pyerf
pyerf.erfinv(0.5) # 0.476936...
pyerf.erf(0.5) # 0.5204998...
pyerf.erfc(0.5) # 0.4795001...
or only a specific function:
from pyerf import erfinv as inverse_error_function
inverse_error_function(0.5) # 0.476936...
and lastly, you can even use import * (but that’s no longer considered
very Pythonic as it pollutes the namespace):
from pyerf import *
pyerf.erfinv(0.5) # 0.476936...
pyerf.erf(0.5) # 0.5204998...
pyerf.erfc(0.5) # 0.4795001...
Changelog¶
See CHANGELOG.md.
pyerf.pyerf¶
This is the main module for PyErf.
-
pyerf.pyerf.erfinv(z)[source]¶ Calculate the inverse error function at point
z.This is a direct port of the SciPy
erfinvfunction, originally written in C.Parameters: z (numeric) – Returns: Return type: float References
- https://en.wikipedia.org/wiki/Error_function#Inverse_functions
- http://functions.wolfram.com/GammaBetaErf/InverseErf/
Examples
>>> round(erfinv(0.1), 12) 0.088855990494 >>> round(erfinv(0.5), 12) 0.476936276204 >>> round(erfinv(-0.5), 12) -0.476936276204 >>> round(erfinv(0.95), 12) 1.38590382435 >>> round(erf(erfinv(0.3)), 3) 0.3 >>> round(erfinv(erf(0.5)), 3) 0.5 >>> erfinv(0) 0 >>> erfinv(1) inf >>> erfinv(-1) -inf