shapelets

Contact: griffin.foster@gmail.com

A python module for fitting and decomposing images (FITS, PNG, JPEG...) into shapelet coefficients, support for Cartesian and polar forms using Hermite and Laguerre polynomials.

Based on the shapelet framework developed in Shapelets: I. A Method for Image Analysis and the IDL shapelet software.

Required Python Modules

• matplotlib
• numpy
• scipy
• astropy
• json

Optional Python Modules

• python-casacore
• PIL

Install

To install the current stable version (0.2) use pip:

pip install shapelets

While developing it is useful to do a developer install:

sudo python setup.py develop

Otherwise, the standard install will work:

sudo python setup.py install  

Usage

The scripts directory contains a number scripts for plotting, decomposing, and fitting.

• plotShapelets.py : plot a grid of shapelet basis functions
• plotImg.py : load a FITS or image file, useful for determing coordinates to apply shapelet decomposition
• plotCoeffs.py : given a shapelet coefficient file, plot the modelled source and coefficients
• solveShapelet.py : given a set of parameters and an image, decompose the image into shapelet coefficients
• fitShapelet.py : fit parameters to minimize the chi^2 difference between shapelet model and image
• insertShapelet.py : insert a shapelet coefficient set into a Measurement Set (requires python-casacore)

Examples

plotShapelets.py -n 4 -p -b 1.0,1.0,0.44

plotImg.py -r 1010,1117,947,1030 ../data/N6251_test.fits
plotImg.py ../data/zen.2455819.69771.uvcRREM_briggs-dirty.fits

solveShapelet.py -r 1010,1117,947,1030 -N 891,1257,600,840 ../data/N6251_test.fits -n 15 -x 49,52 --beta=6.,2.5 --phi=-0.448243 -m cart
solveShapelet.py -r 489,530,489,527 -N 436,561,405,487 ../data/zen.2455819.69771.uvcRREM_briggs-dirty.fits -n 10 -m cart

fitShapelet.py -r 1010,1117,947,1030 -N 891,1257,600,840 ../../data/N6251_test.fits -n 8 -x 49,52 --init_beta=6.,2.5 --init_phi=-0.448243 -m cart
fitShapelet.py -r 489,530,489,527 -N 436,561,405,487 ../../data/zen.2455819.69771.uvcRREM_briggs-dirty.fits -n 8 -x 20.,22. --init_beta=2.788068,2.336974 --init_phi=-1.046530 -m cart -B 10

plotCoeffs.py ../data/solve/N6251_cart.pkl