incorrect initial temperature reported
By comparing the two md.log (2016.4 vs 2018-beta1) I noticed a difference in the evaluation of the initial temperature:
Initial temperature: 600.828 K
Initial temperature: 300.414 K
In the code (md.cpp:780) the temperature is multiplied by a factor of 2, but it was already like this in gmx2016.4.
#3 Updated by Mark Abraham over 1 year ago
Aleksei Iupinov wrote:
Just guessing that it might be related to I87ab08d21a35621735ab3c65fc50af9992120be3 ?
I bisected, diffing the "^Initial temperature" on the log file of complex/nbnxn_rf, and indeed this commit (a15d819d005204b4ccd9e5b40e84166098dbdbc8) introduces the issue
#7 Updated by Berk Hess over 1 year ago
- Status changed from New to Fix uploaded
- Assignee set to Berk Hess
- Target version changed from 2016.5 to 2018
The bug in 2014 is a different one: #2199. That is fixed for 2014.5. This is a new bug.
Szilard, do you mean backport to the 2018 or 2016 release? This should go in 2018, but I pushed up before the branch was created. Now Mark cherry-picked it.
Note that both bugs are only reporting issues, they don't affect any results.