README
  
o Archive
   - Historical data
  
o chrompic_recomb_ord.sh
   - A Linux/Unix shell script to parse chrompic output to extract the phase
     line outputs that have at least a specified number of recombinations.
     The output file contains the phase line on the left (position in line
     indicates affected locus as no spaces in line) and the animal id and
     maternal or paternal on the right (tab delimited).  Currently lacks
     family identity. This script supersedes postchromic.txt which contained
     a bug.
   - Credits: Jill Maddox [jfmaddox@bigpond.com]
  
o crimap_2.503a_inc_doc.tar.gz
   - Improved cri-map source code
   - Credits: Ian Evans, Jill Maddox [jfmaddox@bigpond.com]
  
o crimap_2.504.tgz
   - An updated version of CRI-MAP (v2.504) that resolves some inconsistencies
     between twopoint and other CRI-MAP functionality, also has some additional
     changes that are described in the crimap2_504.doc file that is within the
     crimap2.504.tgz file.  This file should replace the v2.503 on the main part 
     of the site but should leave v2.503 accessible also in case people want to
     continue to use it rather than v2.504. There are slight differences in map
     lengths and likelihood values with v2.504 compared to earlier versions of
     CRI-MAP because recombinations are now treated as integers.
  
   - Compiled with debug codes.
   - Contact: Jill Maddox [jfmaddox@bigpond.com]
  
o crimap_2.504_MacOSX-10.8.zip
   - This program was compiled from "crimap_2.504.tgz" on MacOS 10.8 (mountain
     lion).  The compilation is with the default settings of defs.h and the
     following settings: MAX_NUM_LOCI=11000; MAX_NUM_SWITCHS=1000000;
     MAX_PRODUCT_CALCS=500000000LL; MAX_PRODUCT_CACHE_MEMORY=2000000000LL;
     HIGH_MEMORY_BITS=25. (please check your machine for the specifics for the
     best performance).
  
   - Contact: Paris Veltsos [parisveltsos@gmail.com] (University of St Andrews)
  
o crimap_2.504a.tgz
   - Changes made:
     1. use of float.h rather than values.h in e_chrom.c, e_ped.c, get_like.c
        and screen_o.c.  values.h is now obsolete in and should be replaced 
        by float.h as not all versions of C/C++ will still have values.h. 
        The values.h problem will occur for people using MacBooks and maybe 
        some others. This change should have no effect on functionality.
     2. Makefile with debug turned off for default compiling. Having crimap
        compiled with debug causes it to run more slowly - is only useful if 
        crimap is crashing.
     3. Formating changes in flips.c - a mea culpa as I forgot to change all
        lines to have a spacing of 5 characters, the extra lines that need to
        be adjusted for spacing changes are indicated in crimap2_504a.doc
     4. Flips permitted to have same loci in both ordered_loci and inserted_loci
        lines in par file.
  
   - Compiled without debug codes.
   - Contact: Jill Maddox [jfmaddox@bigpond.com]
  
o crimap_2.507.tgz
   The main change in this version pertains to the implementation of a version 
   of parallelisation using forks and pipes for the get_likelihoods function. 
   This means that flips, all and build will use multiple cpus when more than 
   one order is being tested. Note: this form of parallelisation will not work 
   for Windows but will work for Linux (tested using gcc/g++ versions 4.7.2 and 
   4.1.2) and should work for Unix, Solaris and MacOS.
  
o crimap_2.507_MacOSX-10.10.4.gz
   CRI-MAP 2.507 version compiled for Mac OSX users by Paris Veltsos. Compiled
   using MacOSX 10.10.4 (Yosemite). Note that users should use the command line
   "-n" option to specify the number of cpu's to use as they may not wish to use
   the same number that Paris compiled the program with.

o unmerge_src_310112.tar.gz
   - An improved version of the unmerge program for extracting or deleting loci
     or families from gen files. This version supersedes unmerge_ian.tar.gz
     which had a bug when deleting loci from a gen file. Please note that
     extract works better than delete as the unmerge program still needs to have
     the deleting families option fixed. Whilst one can extract loci and families
     at the same time, one can't combine deletes.
   - Credits: Ian Evans.
   - Contact: Jill Maddox [jfmaddox@bigpond.com]