User Guide

This is how to use Pylira:

from pylira import LIRADeconvolver
from pylira.data import point_source_gauss_psf

data = point_source_gauss_psf()
data["flux_init"] = data["flux"]
deconvolve = LIRADeconvolver(
    alpha_init=np.ones(np.log2(data["counts"].shape[0]).astype(int))
)
result = deconvolve.run(data=data)

The data object is a simple Pythin dict containing the following quantities:

Quantity	Definition
counts	2D Numpy array containing the counts image
psf	2D Numpy array containing an image of the PSF
exposure (optional)	2D Numpy array containing the exposure image
background (optional)	2D Numpy array containing the background / baseline image

From these quantities the predicted number of counts is computed like:

\[N_{Pred} = \mathrm{PSF} \circledast (\mathcal{E} \cdot (F + B))\]

Where \(\mathcal{E}\) is the exposure, \(F\) the deconvovled flux image, \(B\) the background and \(PSF\) the PSF image.