Page 1 of 1

Reproducing Results

Posted: Tue Jul 30, 2019 7:04 am
by moez
Hi,
I recently joined the Forum; sorry if I'm newbie. I tried to see if my question is asked before but I could not find any similar topic in the forum.
I understand that to protect contribution integrity and to prevent reverse engineering, source code of the client application cannot be published. However, as I it is mentioned in the FAQ page of the project home-page, it is possible to develop plug-ins for the client program.
Basically what I want to do is to make my contribution verifiable by third party. To be specific, I want to see if it is possible to record the input (work units), run the simulation (as black-box), and record the output; then later, running the client program in the offline mode, reproducing the results and cross-validating it.
Any guidance would be appreciated a lot. Also, if such features are not implemented yet, I am willing to contribute.
Best,
Moez

Re: Reproducing Results

Posted: Tue Jul 30, 2019 1:13 pm
by JimboPalmer
I would read up on FAHBench to see if it is what you want.

http://fahbench.github.io/overview.html

If it isn't, the code will be a great help writing what you do want. I can't make sense of what you say, it is almost as if you needlessly want to run WUs over and over, instead of doing more science.

Re: Reproducing Results

Posted: Tue Jul 30, 2019 4:18 pm
by bruce
Please visit http://openmm.org/

The analysis code used by FAH is open source