Page 1 of 1

mesfree: MES using chi square free

Posted: 2014.04.24 19:46
by fullerj
I've adapted the MES ensemble search program (original here) to use the chi square free metric from Rambo and Tainer (Nature 2013) as the "fitness" metric in the genetic algorithm, rather than just the reduced chi squared. I think it's at a point where others could give it a shot.

Because chi square free is considerably more computationally intensive, please compile locally to optimize for your system and install the OpenMP libraries to enable multithreading. This is only for linux at this point.

Download:
http://realspace.uchicago.edu/wordpress ... ree.tar.gz
More info:
http://realspace.uchicago.edu/~fullerj@ ... .edu/?p=46

Example usage (where <NS> = the number of nyquist shannon bins and N is the desired subset size):

Code: Select all

./mesfree masterfilelist --threads 4 --population-size 20000 --max-iterations 10000 --num-runs 10 --max-time 14400 --ns <NS> --local-step 15 --output-file mesN.fit --subset-size N > mesN.out