Here I store link to some older general topics. Much of the software methodology descibed in the links below is fairly obsolete, but worth keeping around as a record. I also store here at the end of this doc some notes on my OTW AND OTHER CODES.
This newer generation of routines, compiled under gfortran (specifically that from gcc.4.7.3) is now referred to as lmgf, and both the main and library source codes reside in:
$scohome/codes/fortran/gfortran_compiler Some helpful aliases: mop - go to the main code source directories lop - go to the library directoryThe lmgf acronym stands for LMorpho GFortran. I have revived some of my olde LMORPHO g77 codes that have proven to be useful. Things like XYP and STACKER are still used a lot. In the case of XYP I rely of plotting routines in PGPLOT. This package is becoming more and more difficult to maintain, and so may be abandoned in the not distant future.
In the Fall of 2013, I began writing a new generation of tools, mostly for image processing in astronomical applications. These tools are usually compiled under gfortran, but may eventually be developed in python or other high level languages. The main idea here was to go back to the old unix philosophy of "One Thing Well" (hence the otw). In these codes, I usually take as input images with the name "in.fits", and if I am producing an output image it will be named "out.fits". In nearly every set of source code (i.e. name/name.f) I have a shell script (name/name.sh) that will run the code from the command line with no user interaction. The main and library source codes for OTW reside in:
$scohome/codes/fortran/gfortran_compiler/otw/ Some helpful aliases: mopo - go to the main code source directories lopo - go to the library directoryTo insure that each OTW code has a run script that is properly installed, I wrote the script "botw" that transfers the compiled executable and the run script to the proper bin directory. An example of running this is show below:
% pwd /home/sco/sco/codes/fortran/gfortran_compiler/otw/src/thresh % botw thresh gfortran -O -ffixed-line-length-none -fno-automatic -c -o thresh.o thresh.f gfortran -o thresh -O -ffixed-line-length-none -fno-automatic thresh.o /home/sco/sco/codes/fortran/gfortran_compiler/libs/LIN/OTWLIB.a -L/usr/lib/x86_64-linux-gnu -lX11 a makefile makes T thresh thresh.f thresh.log thresh.o thresh.sh I have built and installed thresh
Finally, I have begun to keep test images and data files (small ones!) in a single directory structure: Test_Data_for_Codes.
% pwd /home/sco/sco/projects/Test_Data_for_Codes % ls README.Test_Data_for_Codes T_images/ T_runs/ Some helpful aliases: got - go to data directories for code testing gotr - go to data directories for code testing runsThe idea is that I keep run scripts and instruction files in the T_runs subdirectory. The scripts might pull test images from T_images, but the actually run locations have ./S directory where scripts and things are stored. The run locations (like /Test_Data_for_Codes/T_runs/contours_1/ex01) where the code is run should be considered very volatile. The data files and output results can be easily reproduced without keeping everything around in each run directory.
The OTW scripts, in particualry, are designed to one simple task. To perform complex tasks that are a combination of these simple jobs I have been composing run scripts. I will document some of these scripts in the table above, and to make them easier to spot I will print the script names using purple text. Most of these scripts reside in my otw bin directory. As a reminder, the path for this directory on my home office ubuntu machine is:
% pwd /home/sco/sco/bin/binc/otw % ls cef* edger_sqt* find1.sh* fourcon.sh* imcop.sh* scc.sh* thresh.sh* cef.sh* edger_sqt.sh* fourcon* imcop* scc* thresh*