astro-bifrost 2.0.3

Correct, normalize, coadd, and stack spectra together.

BIFRÖST

Black hole Investigations of Forbidden-line Radiation in Optical SpecTra

This code aims to connect galaxy spectra to coronal line detections, just as the burning rainbow bridge it is named after connects Asgard and Midgard, by providing a means of searching for them via a framework for analyzing spectra with simple tests and a stacking procedure.

Installation

After cloning the repository, you can install locally via pip: pip install .
Or alternatively, python setup.py install All of the dependencies are listed in the requirements.txt file, and should be installed automatically. These dependencies are:

• numpy < 1.21, >= 1.17
• numba >= 0.53.1
• pandas >= 1.3.2
• tqdm >= 4.61.2
• matplotlib >= 3.4.2
• plotly >= 4.14.3
• astropy >= 4.2.1
• astroquery >= 0.4.2
• spectres >= 2.1.1
• scipy >= 1.7.0
• joblib >= 1.0.1
• PyAstronomy >= 0.16.0
• kaleido >= 0.2.1

I can't guarantee this package will work for any versions of these requirements outside the suggestions above.

Documentation Summary

Most of the functions of this module are built around three core objects: Spectrum, Spectra, and Stack. Please see the source code for more detailed docstrings for each class and method.

Spectrum

The Spectrum class stores a single object's spectrum, including an array of wavelengths, fluxes, and 1-sigma errors. Information about redshift, coordinates, and extinction can also be stored and (if available) used to apply corrections to the spectrum. The basic constructor is:
class Spectrum(wave, flux, error, redshift=None, velocity=None, ra=None, dec=None, ebv=None, name='Generic', output_path=None, **data)

There are also constructors available for creating an object directly from an SDSS-formatted FITS file: Spectrum.from_fits, and from completely simulated data: Spectrum.simulated.

As a side note, basic arithmetic operators (+, -, *, /) may be performed between Spectrum objects. Doing so will create a new Spectrum object with the flux combined as expected, and the error will be propagated as well. This is only possible if the two objects being added have identical wavelength arrays and are both corrected, and they must also either both or neither be normalized.

Spectra

The Spectra object is a child of both Spectrum and dict. It is a convenient way of storing many Spectrum objects together in a dictionary, with methods for correcting and plotting each Spectrum in the dictionary. Also as a convenience, the object can be indexed like a list or a dictionary. i.e. if the first element in the dictionary has a key 'Spectrum A', it can be obtained either using the key 'Spectrum A', or the index 0. There is also a method called to_numpy for converting any attributes into numpy arrays. Its basic constructor is no different from a typical dictionary:
class Spectra()
and items can be added with the add_spec method: Spectra.add_spec(SpectrumA)

Stack

The main attraction, the Stack object, is a child of the Spectra object, thus making it also dictionary-like. This class has methods for stacking all the spectra inside it together, binning the spectra and then stacking each bin, calculating line flux ratios, and many plotting methods.

To perform the stacking procedure, (which corrects all spectra in the Stack for redshift and extinction, normalizes them, resamples their fluxes over a uniform wavelength grid, and then coadds them using a weighted mean, with inverse variances used as weights), one must call the Stack object instance directly, i.e. if we have two Spectrum objects A and B:

# Create the stack object
stack = bifrost.Stack()
# Create spectrum objects
A = bifrost.Spectrum(wave1, flux1, error1, redshift1)
B = bifrost.Spectrum(wave2, flux2, error2, redshift2)
# Append Spectrum objects to the stack
stack.add_spec(A)
stack.add_spec(B)
# Call the stacking procedure with the default arguments
stack()

Examples

See the example files in the examples folder to get a sense of how to use the code. Feel free to also run the unit tests located in unit_test.py.