There is a newer version of the record available.

Published June 28, 2023 | Version v1.0.2
Software Open

aurelio-amerio/gPCS: v1.0.2

Creators

Description

gPCS: Gamma-ray Photon-Counts Statistics

This repository contains the code for the paper Deepening gamma-ray point-source catalogues with sub-threshold information.

We provide our results in the form of a precomputed FITS, as well as python package which can be used to read the data as numpy arrays, as well as export a similar FITS table.

Installation

This package can easily be installed through pip:

pip install gPCS

Example usage

These examples can be run online on colab.

This package can be used to obtain the firing pixels, either given a chosen TS_star or by fixing a quality factor (QF) and significance level (alpha).

import numpy as np
from gPCS import gPCS

# specify manually a TS_star
TS_star = 36
pixel_firing = gPCS.get_firing_pixels(TS_star, filter=False)
print(len(pixel_firing))

# Compute the TS_star from a chosen QF and alpha
QF = 0.5
alpha = 0.05
TS_star = gPCS.get_TS_from_QF(QF, alpha=alpha)
pixel_firing = gPCS.get_firing_pixels(TS_star, filter=False)
print(len(pixel_firing))

We can get the TS of the firing pixels:

TS_star=36
firing_pixels = gPCS.get_firing_pixels(TS_star, filter=False) 
TS_ranking = gPCS.TS_map_Fermi[firing_pixels]

It is easy to obtain the galactic coordinates of the firing pixels using healpy:

import healpy as hp
lon, lat = hp.pix2ang(NSIDE, firing_pixels, lonlat=True) # lon lat in degrees

And we can compute the QF and QF range of the firing pixels, given alpha:

# obtain the QF using all the simulations
QF = gPCS.get_QF_from_TS(TS_ranking, alpha=alpha)

# compute the mean and std of the QF using batches of simulations
mean_QF, std_QF = gPCS.get_QF_ranges_from_TS(TS_ranking, alpha=alpha, 
                            batches=100, batch_size=3000)

# we can obtain the QF range for the firing pixels using the mean and std
QF_min = mean_QF - std_QF
QF_max = mean_QF + std_QF

If desired, we provide some simple functions to filter the firing pixels, for example by removing pixels that are firing for the simulated 4FGL K map (see paper):

# filter out the firing pixels for the 4FGL K map
pixel_firing = gPCS.get_firing_pixels(TS_star, filter=True)
# If conservative is True, after filtering the pixels in K, 
# the routine will also filter the pixels in the 1 pixel neighborhood of the pixels in K.
pixel_firing = gPCS.get_firing_pixels(TS_star, filter=True, conservative=True)
# If deg is specified, the routine will also filter the pixels 
# in the disc of radius deg centered on the centroid of the 4FGL catalog sources.
pixel_firing = gPCS.get_firing_pixels(TS_star, filter=True, conservative=True, deg=0.5)

For ease of use, we provide the TS maps both for Fermi and the simualted 4FGL map (K) as numpy attays. The maps are computed at nside=512 and are stored in the following variables:

gPCS.TS_map_Fermi
gPCS.TS_map_4FGL

In order to compute the 4FGL K map, we use employ the gll_psc_v30 catalog, available from the Fermi LAT colalboration website.

Exporting the results

We provide a simple function to export the results in the form of a FITS table, which can be used to reproduce the results of the paper.

export_fits_table(filename, QF, alpha, overwrite=False, **kwargs)

export_fits_table accepts the following arguments:

  • filename : name of the output FITS file
  • QF : quality factor
  • alpha : significance level (can be an array, and the supported values are 0.01, 0.05, 0.1)
  • overwrite : if True, the routine will overwrite the output file if it already exists
  • **kwargs : additional arguments to be passed to get_firing_pixels.

export_fits_table will create a FITS table with the following columns:

  • pixel : pixel index
  • TS : TS value
  • QF_best : QF value obtained by considering all the simulations
  • QF_min : lower bound of the QF range
  • QF_max : upper bound of the QF range

In order to export the FITS table available in the examples folder, we can run the command:

gPCS.export_fits_table(filename="firing_pixels.fits", QF=0.50, alpha=[0.01, 0.05, 0.1])

List of functions

  • get_QF_from_TS(TS, alpha): computes the quality factor from a given TS and alpha

  • get_QF_ranges_from_TS(TS, alpha, batches=100, batch_size=3000): computes the mean and std of the QF from a given TS and alpha, using batches of simulations

  • get_TS_from_QF(QF, alpha): computes the TS from a given QF and alpha

  • get_firing_pixels(TS_lim, filter=False, conservative=False, deg=None): computes the firing pixels for a given TS_lim.

  • export_fits_table(filename, QF, alpha, overwrite=False, **kwargs): exports the results in the form of a FITS table.

For more information about what each function does, please refer to the docstrings and help of each function.

License

This code is released under the Zlib license. See the LICENSE file for more information.

Files

aurelio-amerio/gPCS-v1.0.2.zip

Files (38.3 MB)

Name Size Download all
md5:0e98fb89ef8ecf291936146b1f1d907b
38.3 MB Preview Download

Additional details

Related works