TESTSUITE FOR MIGRATE

This README covers testing of migrate using different datasets
using a Baysian or ML inference. The datasets all contain 3 
populations and the analyses use connection matrix that exercises 
symmetric and asymmetric immigration rates and sets some 
migration routes to zero, using heaating to calculate thermodynamic 
integrated marginal likelihoods and multiple replicates. I believe 
that this exercise that the migration model exercises most 
of migrate's code. 

The datasets are
- dna data
- hapmap data
- microsatellite data
- allele data
The datasets are real datasets and no simulated.

the python program rt takes these options either alone or in combinations
-single
-thread
-parallel
--nocompile [compile is the default]
--cores=X   [user specified number of cores]

a call 
rt -single
will run the 4 datasets for 2 strategies and should generate 8 pdf files with results,
rt -thread -parallel
will run 8 datasets using the thread and the parallel migrate versions.
The script compiles the program versions, threfore it needs to be run on a Mac or Linux
and access to the source. This direcotyr needs to be placed in the source directory.

A set will take considerable time (several hours) on a standard machine.

If you run this suite for all 3 program versions then you create 24 outfiles.
If you run this testsuite, let me know when it fails and generates fewer than 
8 PDF outfiles for each version.
 
A python program "reporter" delivers times in seconds for the different scenarios,
if you run this testsuite, send me the output of reporter
and I will update my website to give some impression how fast migrate can run
on clusters and desktop machines.

A shell script "cleaner" removes all the generate files.

Here an example from my 2-core macbook pro using
rt -single -parallel | sort

parallel  bayes dna 318
parallel  bayes hap 421
parallel  bayes iam 38
parallel  bayes smm 500
parallel  ml dna 386
parallel  ml hap 590
parallel  ml iam 22
parallel  ml smm 92
single  bayes dna 287
single  bayes hap 311
single  bayes iam 35
single  bayes smm 317
single  ml dna 432
single  ml hap 430
single  ml iam 21
single  ml smm 103

