make.deblend.weight.maps: Deblend Photometric Objects

View source: R/make.deblend.weight.maps.R

make.deblend.weight.mapsR Documentation

Deblend Photometric Objects

Description

__This function is not designed to be called directly, but rather is an internal function to measure.fluxes__

Deblends objects which are overlapping in a photometric image. Procedure is called by flux.measurements, part of the measure.fluxes routine. The procedure takes individual weighted object apertures, and determines what proportion of the blended flux should belong to the object in question. It performs this analysis on all apertures from the input catalogue.

Usage

make.deblend.weight.maps(outenv=parent.env(environment()),
env=NULL,subs=NULL)

Arguments

env

environment; the environment containing the parameters used in ‘measure.fluxes’

outenv

environment; the environment where the returned values are to be stored. If not provided, outenv defaults to the input env.

subs

numeric length(k) vector; the subset of apertures to calculate deblending for. Default NULL is treated as all (i.e. 1:ngal).

Details

The most relevant parameters to the execution of this function are the weighted & convolved aperture stamps (wsfa), and the weighted & convolved full aperture mask (wfa):

- wsfa: numeric length (m) list of (n,n) arrays; a length (m) list of (n,n) weighted & convolved object aperture stamps - wfa: numeric (p,q) array; full mask of all weighted & convolved object apertures, placed in their appropriate locations, coadded

The deblended weightmaps are determined by what fraction of each aperture contributes to the full mask model at each pixel in the objects stamp. I.e. Deblended weightmap @ (i,j)th pixel of an aperture stamp = wsfa[i,j]/wfa[n,m], where (n,m) is the corresponding location of (i,j) in the full mask.

Value

numeric length (m) list of (n,n) arrays; a length (m) list of (n,n) weighted, deblended, object aperture stamps

Author(s)

Angus H Wright ICRAR angus.wright@icrar.org

See Also

measure.fluxes, flux.measurements


AngusWright/LAMBDAR documentation built on May 12, 2022, 1:49 a.m.