UNDER DEVELOPMENT

The software package is currently being dramatically improved and updated.
This is a (very) old version, potentially with bugs.

Shapelets web page	Shapelets IDL code	Installation	Help pages	Simulated images
			List of routines
			Instrutions for general code
			Instructions for image simulation code

Help for "simage_assemble_image" routine:

Takes objects from a new shapelet catalogue and scatters them about an image, without noise.

This routine takes a shapelet catalogue, and lays down the objects within a blank image array. It ouputs the image as image.fits, and a faked SExtractor file image.sex containing each object's (known) input position, size, magnitude, ellipticity and star/galaxy classification. The image's header also contains a lot of information concerning the number and type of the simulated objects. This routine can be called from a shell prompt via the generate_simage script.

Note that the input shapelet catalogue could be either a catalogue containing morphed, new objects created by simage_generate_shapecat.pro, or the catalogue of a real image. I recreated both HDFs exactly, as a test of the method. To do this, ensure that the /RPOS flag is not used. This (re-)randomises the positions and orientations of objects in the input catalogue.

The new image can contain galaxies created from either a full shapelet model (c.f. simage_make_shapelet_object.pro), or just an analytic radial profile (c.f. simage_make_analytic_object.pro), with elliptical isophotes. I've been using this option to test shapelet-based shear measurement methods without biases created by measuring shear from a shapelet model by using a shapelet model. I don't know if it will be of any use, but I've left it in just in case. Weak gravitational lensing shear can be added to the simulated image, at a constant value across the field, with the input parameter gamma.

The subroutine simage_locate_object adds a postage stamp image containing one object to the overall image, making sure that the postage stamp is truncated at the same places as the edges of the full image. More information about the subroutines is available in their individual headers, within the source code for this routine. The subroutine simage_star_magnitude_distribution generates a random flux level for stars in the image. The stars will follow the stellar luminosity function measured at the galactic poles and tabulated in Allen's astrophysical quantities. Note that the overall number density of stars (and galaxies) is a user input, as is the overall size of the simulated image. Only the slope of the number counts is guaranteed to be correct (but in the case of the galaxies, depends upon the slope of the number counts in the input shapelet catalogue). The default values create an image with the same number of pixels and object number density as the Hubble Deep Fields.

See also the header below, which has been extracted from the source code for this routine.

; NAME: ; SIM_ASSEMBLE ; ; CATEGORY: ; Shapelets image simulation. ; ; PURPOSE: ; Collects newly synthesised objects from sim_create.pro, rotates them ; by a random amount, and scatters them into pixels to make a simulated ; image. Background levels and noise are then be added using ; simage_add_noise.pro. ; ; This routine can also create a simulated image containing objects with ; analytic (e.g. de Vaucouleurs) profiles with the same size, magnitude ; and ellipticity distributions as a shapelet catalogue (e.g. for the HDF). ; The shapelet coefficients are not used for any purpose other than the ; determination of these distributions (which may also be entered by hand). ; The point is to use a non-shapelet method to create simulations, which ; can themselves be then used to test shapelet-based shape-measurement ; techniques. ; ; INPUTS: ; cat - Synthesised catalogue of shapelet objects, eg created by new_hdf ; file - Filename for output w/o extension, eg "newimage" ; ; OPTIONAL INPUTS: ; [seed] - Seed for random number generator. ; [noise] - Exposure time to pass to simage_add_noise.pro ; [gamma] - Weak lensing shear on all galaxies [gamma_1,gamma_2]. ; [psf] - Cartesian decomp structure representing PSF. Will then do ; convolvution automatically. Stars are just this PSF. ; [inst] - Post-smear shear due to geometric distortions in telescope. ; [pixsize] - Pixel scale [arcsec]. Default is 0.04". ; [n_pix] - Number of pixels in final image [x,y]. ; [n_gal] - Number of galaxies in final image. ; [n_star] - Number of stars in final image (currently all same magnitude). ; [dither] - Dither offset of exposure [x,y]. (Currently a simple linear shift ; which is the same everywhere and should be a fraction of a pixel). ; [parameter] - Use parameterised models of galaxy shapes rather than shapelet ; coefficients. ; 1: de Vaucouleurs r^(1/4) profiles ; 2: Exponential discs ; 3: Elliptical Gaussian blobs ; [sigmae] - rms galaxy ellipticity (before shear) for use with ; parameterised galaxy shapes. ; ; KEYWORD PARAMETERS: ; /RPOS - (Re-)randomise positions and PAs of galaxies? 1=YES, 0=NO ; ; OUTPUTS: ; Image, written to disc. ; This image is also (optionally) returned to the variable "image", with ; its FITS header in the variable "header". ; ; MODIFICATION HISTORY: ; Oct 2003 - Centroid bug fixing by RM. ; Jul 2003 - Input file in SExtractor format added by RM. ; Jun 2003 - Broken into separate modules by RM. ; Apr 2003 - Bugs fixed and analytic profiles forms added by RM. ; Feb 2002 - Written by Richard Massey.

Return to the shapelets web page or the code help menu.

Last modified on 02nd May 2008 by Richard Massey.