spectrum

class pyspeckit.spectrum.Spectrum(filename=None, filetype=None, xarr=None, data=None, error=None, header=None, doplot=False, maskdata=True, plotkwargs={}, xarrkwargs={}, **kwargs)

Bases: object

The core class for the spectroscopic toolkit. Contains the data and error arrays along with wavelength / frequency / velocity information in various formats.

Contains functions / classes to:

-read and write FITS-compliant spectroscopic data sets

-read fits binaries? (not implemented)

-plot a spectrum -fit a spectrum both interactively and non-interactively

-with gaussian, voigt, lorentzian, and multiple (gvl) profiles

-fit a continuum “baseline” to a selected region with an n-order

polynomial

-perform fourier transforms and operations in fourier space on a spectrum (not implemented)

Example usage:

# automatically load a “compliant” (linear X-axis) FITS file

spec = pyspeckit.Spectrum(‘test.fits’)

# plot the spectrum

spec.plotter()

__init__ Initialize the Spectrum. Accepts files in the following formats:

• .fits
• .txt
• .hdf5

Must either pass in a filename or ALL of xarr, data, and header, plus optionally error.

doplot - if specified, will generate a plot when created

kwargs are passed to the reader, not the plotter

copy(deep=True)

Create a copy of the spectrum with its own plotter, fitter, etc. Useful for, e.g., comparing smoothed to unsmoothed data

crop(x1, x2, units=None, **kwargs)

Replace the current spectrum with a subset from x1 to x2 in current units

Fixes CRPIX1 and baseline and model spectra to match cropped data spectrum

getlines(linetype='radio', **kwargs)

Access a registered database of spectral lines. Will add an attribute with the name linetype, which then has properties defined by the speclines module (most likely, a table and a “show” function to display the lines)

interpnans(spec)

Interpolate over NAN values, replacing them with their neighbors...

measure(z=None, d=None, fluxnorm=None, miscline=None, misctol=None, ignore=None, derive=True, **kwargs)

Initialize the measurements class - only do this after you have run a fitter otherwise pyspeckit will be angry!

moments(unit='km/s', **kwargs)

Return the moments of the spectrum. In order to assure that the 1st

and 2nd moments are meaningful, a ‘default’ unit is set. If unit is not set, will use current unit.

Returns (height, amplitude, x, width_x)

the gaussian parameters of a 1D distribution by calculating its moments. Depending on the input parameters, will only output a subset of the above. “height” is the background level “amplitude” is the maximum (or minimum) of the data after background subtraction “x” is the first moment “width_x” is the second moment

If using masked arrays, pass estimator=np.ma.median ‘estimator’ is used to measure the background level (height)

negamp can be used to force the peak negative (True), positive (False), or it will be “autodetected” (negamp=None)

“nsigcut” - try to estimate the noise and only use data above/below nsigma

above the noise. Estimate the noise from the data unless passed as a keyword

Theory: From first principles (in the absence of noise): integral(gaussian) = sqrt(2*pi*sigma^2) * amp sigma = integral / amp / sqrt(2*pi)

in the presence of noise, this gets much more complicated

parse_hdf5_header(hdr)

HDF5 reader will create a hdr dictionary from HDF5 dataset attributes if they exist. This routine will convert that dict to a pyfits header instance.

parse_header(hdr, specname=None)

Parse parameters from a .fits header into required spectrum structure parameters

This should be moved to the FITSSpectrum subclass when that is available

parse_text_header(Table)

Grab relevant parameters from a table header (xaxis type, etc)

This function should only exist for Spectrum objects created from .txt or other atpy table type objects

shape()

Return the data shape

slice(start=None, stop=None, units='pixel', copy=True, preserve_fits=False)

Slicing the spectrum

start: numpy.float or int

start of slice

stop: numpy.float or int

stop of slice

units: str

allowed values are any supported physical unit, ‘pixel’

smooth(smooth, **kwargs)

Smooth the spectrum by factor “smooth”.

sm.smooth doc:

Smooth and downsample the data array

smooth [ float ]

Number of pixels to smooth by

smoothtype [ ‘gaussian’,’hanning’, or ‘boxcar’ ]

type of smoothing kernel to use

downsample [ bool ]

Downsample the data?

downsample_factor [ int ]

Downsample by the smoothing factor, or something else?

convmode [ ‘full’,’valid’,’same’ ]

see numpy.convolve. ‘same’ returns an array of the same length as ‘data’ (assuming data is larger than the kernel)

stats(statrange=(), interactive=False)

Return some statistical measures in a dictionary (somewhat self-explanatory)

range - X-range over which to perform measures interactive - specify range interactively in plotter

write(filename, type=None, **kwargs)

Write the spectrum to a file. The available file types are listed in spectrum.writers.writers

type - what type of file to write to? If not specified, will attempt to determine type from suffix

class pyspeckit.spectrum.Spectra(speclist, xunits='GHz', **kwargs)

Bases: pyspeckit.spectrum.classes.Spectrum

A list of individual Spectrum objects. Intended to be used for concatenating different wavelength observations of the SAME OBJECT. Can be operated on just like any Spectrum object, incuding fitting. Useful for fitting multiple lines on non-continguous axes simultaneously. Be wary of plotting these though...

Can be indexed like python lists.

X array is forcibly sorted in increasing order

fiteach(**kwargs)

Fit each spectrum within the Spectra object

ploteach(xunits=None, inherit_fit=False, plot_fit=True, plotfitkwargs={}, **plotkwargs)

Plot each spectrum in its own window inherit_fit - if specified, will grab the fitter & fitter properties from Spectra

smooth(smooth, **kwargs)

Smooth the spectrum by factor “smooth”. Options are defined in sm.smooth

because ‘Spectra’ does not have a header attribute, don’t do anything to it...

class pyspeckit.spectrum.ObsBlock(speclist, xtype='frequency', xarr=None, force=False, **kwargs)

Bases: pyspeckit.spectrum.classes.Spectra

An Observation Block

Consists of multiple spectra with a shared X-axis. Intended to hold groups of observations of the same object in the same setup for later averaging.

ObsBlocks can be indexed like python lists.

average(weight=None, inverse_weight=False, error='erravgrtn')

Average all scans in an ObsBlock. Returns a single Spectrum object

weight - a header keyword to weight by error - estimate the error spectrum. Can be:

‘scanrms’ - the standard deviation of each pixel across all scans ‘erravg’ - the average of all input error spectra ‘erravgrtn’ - the average of all input error spectra divided by sqrt(n_obs)

smooth(smooth, **kwargs)

Smooth the spectrum by factor “smooth”. Options are defined in sm.smooth

pyspeckit.spectrum.register_reader(filetype, function, suffix, default=False)

Register a reader function.

Required Arguments:

filetype: [ string ]

The file type name

function: [ function ]

The reader function. Should take a filename as input and return an X-axis object (see units.py), a spectrum, an error spectrum (initialize it to 0’s if empty), and a pyfits header instance

suffix: [ int ]

What suffix should the file have?

Optional Keyword Arguments:

pyspeckit.spectrum.register_writer(filetype, function, suffix, default=False)

Register a writer function.

Required Arguments:

filetype: [ string ]

The file type name

function: [ function ]

The writer function. Will be an attribute of Spectrum object, and called as spectrum.Spectrum.write_hdf5(), for example.

suffix: [ int ]

What suffix should the file have?

Optional Keyword Arguments: