This repository contains tools developed to reduce a set of DECam images acquired to map SII and Ha emission in the Magellanic Clouds
To use these tools this directory should be placed both in PATH and PYTHONPATH
If one has placed kred in one's home directory, the following should be added to ones .bashrc or .bash_profile
export KRED=$HOME/kred/
PATH=$PATH:$KRED/py_progs/
PYTHONPATH=$PYTHONPATH:$KRED/py_progs/There is an Jupyter script that performs very basic tests of the installation in the exmaples directory.
The tools require a set of relatively commmon modules, such as astropy, and normally are run from within an appropriate conda enviromnment.
The version of astropy that is used is relatively recent
If you encounter problems, one should create a specific conda enviromment with something like the following commands:
conda create --name kred scipy
conda install astropy
conda install matplotliband run in this enviroment which should have versions of these three packages greater than or equal to:
scipy 1.10.1
astropy 5.1
numpy 1.25.0If one wishes to use the ancillary routines that involve the GAIA data one must also have the following installed
photutils
astroquery
GaiaXPyThe routines assume a specific directory structure to one of our own. At present the new structure presumes:
The programs are all run from a toplevel directory. Underneath this directory there must be:
Additional directories will be created as various programs are run.
To get help on the options of the various routines, one should type the program name, with a -h option.
Several of the programs have a flag -bsub that change the data on which the operate. The first discussion of each routine assumes the -bsub flag is NOT invoked so that users can get a clearer idea of the flow. The -bsub option will be explained later
To process the data without attempting to improve over the background options in MefPrep, the programs should be run in the following order:
(MefCheck.py - This should not normally be necessary, but it will check if there are problems with the headers in the individual Mef files. It creates a table called _mef_qual.tab which indicates which Mef files are problematic, but is not used by any of the downstream programs.)
MefSum.py - This summarizes the mef files in a field. If it does not already exist this program creates a directory that is created called Summary and writes two astropy tables that contain the summary information
Note that MefSum.py -all wil attempt to create a summary of all of the files that one has in the DECam_MEF directory. This routine is parallelized; each thread handles an individual mef file. (In a previous version, the routine was parallelized so that each thread handled a different field; the new version is faster for processing a single field, but not if you are processing a large number of fields.) The program does not check to see if certain Fields have already been analyzed, and so if one adds a single Field one should not use the -all option.
Note - In the current version of the program the most important statistic is produced is the Mode, which is nominally what is normall taken for the background. The mean, median, and standard deviation, which are also calculated are not the sigma clipped versions. There is an option to create sigma clipped versions of the median, and mean, but this takes about a factor of 3 times longer to run.)
Note - this routine has switches that allow one to subtract background or not from the original images. The default currently is to subtract a single background from all of the CCDS associated with a given exposure. The default is to subtract the median value of the mode for the individual detectors.
Also, like MefSum.py one can run this on all of the fields with the -all switch.
Note - This routine uses the config file. If one wanted to create data that was for a different region, then one would need to point this to a different configuration file. Note though, that at present each field is separate.
Note - At present this routine does not use the field centers defined in the .config file, but takes the center of all of the images in a tile to be the center of the swarp field.
When run without the -bsub flag, the routine creates directories in DECam_SWARP and sets up the inputs so the files which will be 'swarp'ed come from the subdirctories of DECam_CCD.
Swarp.py - The purpose of Swarp.py is to run all of the commands that have been created with SetupSwarp.
SwarpEval.py - This routine creates a set of evaluation figures for the Swarped images. The scaling is intended to highlight flaws. The images for a field are stored in the directory DECam_SWARP/field/eval where field is the field name, e.g. LMC_c42
CleanStars.py - This routines attempts to create continuum subtracted images by first creating emission line fre images of the N708 and r-band images and then subtracting these from the emission line images. The results are stored in the directory DECam_SUB/field/tile
All of the scripts in this sequence should be run from the toplevel directory for the data reduction.
The various scripts above allow one to produce swarped imaged for each tile, but they do not include any attempt to match backgrounds between images.
In order to improve backrouund matching one should run the scripts below. They depend on MefPrep having been run on a field, BUT the other steps in the standard processing are unnecessary.
The routines should be run in the following order.
FindOverlaps.py uses information, namely the RAs and DECs of the corners of the indiviual CCDs, contained in the headers of the images to identify what images overlap with others. The routine assumes that the only overlaps we are concerned with are those between images with the same filter and exposure time.
BackPrep.py creates a directory structure DECam_BACK parallel to the other directory structures for storing images created for the purpose estimating background in the overlap regions of images. BackPrep also creates Swarp commands for projecting all of the images (which have gone through the PrepFiles stage) onto the same WCS. This WCS has larger pixels than the original WCS in order to keep the files sizes plausible. When run with the -run option BackPrep then Backprep will run Swarp to create these images.
Note - The files created by BackPrep are large. If one runs BackPrep on all of LMC_c42, for example, about 420 GB of images will be created.
Note _ Optionally, with the -rm option, BackStats removes the files created by BackPrep, to retain diskspace, since BackPrep can generate up to a half a TB of data.
BackCalc.py uses the differences calculated with BackStats between images to determine an optimal set of offsets to add or subtract from the images to produce better image matching.
BackSub.py uses the results of BackCalc to create new images with the backgounds values produced by BalCalc.py subtracted. The outputs appear in a directories, with names of DECam_Prep2/field/tile.
SwarpSetup.py is run (again) but with the -bsub flag in order to create directories of the form DECam_Swarp/field/tile_b and commands in these directories that point to the files cerated by BackSub.py
Swarp.py is run again, but with the -bsub flag to run the comamnds from SetupSwarp.py
SwarpEval.py is run to make images (pngs) of for both types of processing.
CleanStars.py is run with the bsub option. It now uses data in the DECam_SWARP2 directory, and it writes the data to the DECam_SUB2 directory
The progams generally write information to a log file, one log file for each field. The file is initialised by running MefSum, and then each program has the ability to write to it.
When trying to process a significant amount of data, it is useful to carry out the first two steps together, e.g.
MEFSum.py -np 8 LMC_c42
MefPrep.py -np 8 LMC_c42on the field or fields one wants to analyze.
At that point one can create a file to process either a single tile, or a field completely. As an example one could then process all of LMC_c42, with the following set of commands:
SetupTile.py -all LMC_c42
SwarpSetup.py -all LMC_c42
# Optional, if one wants to see swarps based on the original backgourns
Swarp.py -all LMC_c42
SwarpEval.py -all DECam_SWARP LMC_c42
# End optional section
FindOverlaps.py -all LMC_c42
BackPrep.py -all -run LMC_c42
BackStats.py -all -np 8 -rm LMC_c42
BackCalc.py -all LMC_c42
BackSub.py -all -np 8 LMC_c42
SwarpSetup.py -all -bsub LMC_c42
Swarp.py -all -bsub LMC_c42 T02
SwarpEval.py -all DECam_SWARP2 LMC_c42
CleanStars.py -all LMC_c42
SwarpEval.py -all DECam_SUB2 LMC_c42Creating a command file and commenting out the sections that one has completed is a sensible approch to precesing.
Note that the first 4 commands here are not normally necessary and are only useful if one wants to see the immprovement from a more sophistcated image mataching of backgrounds
Two other programs are currently available. Both make use of astropy tables with (minimally) the columns: Source_name,RA,Dec where RA and Dec must be in degrees. (Examples can be found in the config folder, for LMC and SMC SNRs.)
Cutout.py can be used to created cutout images at the positions of a set of objects. Depending on the command line options, one can create cutouts of all the fits images in a particular directory or in all of the images in that directory or any of its subdirectories.
XSnap.py, which is similar to Cutout.py in may respects, also creates cutout positions for a set of objects defined in a "masterfile". In this case, the master file defines not only the center position for the cutout, but also a circular or elliptical shape for each object and overplots that on the image.
SetupSpecial.py can be used to create 'tiles' at an arbitrary set of positions, which can then be processed in the standard manner. Depending on the command line switches, the program will attempt to use only data from specific fields, or all of the data that exists on one's machine.
GetGaiaCat.py is a routine that retrieves catalog data from the GAIA archive for a filed or some tiles in a field
PhotCompare.py is a routine that identifies stars in the DECam images, does simple aperture photometry on them, and then compares the derived magnitudes to magnitudes of GAIA stars in the field.
ZeroPoint.py is a routine that takes a subset of the objects identified with PhotCompare.py and retrieves the low resolution GAIA spectra, and calculates the conversion between DN for the Ha and [SII] filters to flux in ergs/s.
MakeReg.py is a simpple utility routine which creates a set of region files with one region for each fits file with a unque filter and exposure in a directory. The region for each file is defined by the corners of the image, as defined in the header. It is intended to make it easy to overlay on, for example, an MCELS image, the areas covered by a set of exposures.