RE: "-nan" free energy from WHAM analysis

From: Vermaas, Joshua (Joshua.Vermaas_at_nrel.gov)
Date: Mon Jan 28 2019 - 09:37:24 CST

Hi Jun,

I assume you are talking about Alan Grossfield's implementation of WHAM, right? I usually only see nan values when there are empty windows, which might be a result of being too generous with hist_min or hist_max, or having so many bins that a few intermediate ones end up with no samples. If this happens, the nans start propagating rather quickly, and pollute the whole free energy profile.

Unrelated: you may want to try more restrictive tolerances. 1e-6 is typically what I use, and doesn't take that much longer than 1e-4.

-Josh

On 2019-01-27 11:18:14-07:00 owner-namd-l_at_ks.uiuc.edu wrote:

Dear all,
-nan is showing in the output files instead of numerical values after calculating free energy using WHAM. I checked the histograms are appreciably overlapping. I used the same force constant value that I used for umbrella sampling simulations in NAMD. And there was no error messages as well.
The command line used was,
"wham 12 50 180 0.0001 300.0 0 metafile free.pmf"
And my metafile contains,
./output_umbrella/0/win0.colvars.traj 50 10
./output_umbrella/1/win1.colvars.traj 49 10
./output_umbrella/2/win2.colvars.traj 48 10
./output_umbrella/3/win3.colvars.traj 47 10
./output_umbrella/4/win4.colvars.traj 46 10
./output_umbrella/5/win5.colvars.traj 45 10
./output_umbrella/6/win6.colvars.traj 44 10
./output_umbrella/7/win7.colvars.traj 43 10
./output_umbrella/8/win8.colvars.traj 42 10
./output_umbrella/9/win9.colvars.traj 41 10
./output_umbrella/10/win10.colvars.traj 40 10
./output_umbrella/11/win11.colvars.traj 39 10
./output_umbrella/12/win12.colvars.traj 38 10
./output_umbrella/13/win13.colvars.traj 37 10
./output_umbrella/14/win14.colvars.traj 36 10
./output_umbrella/15/win15.colvars.traj 35 10
./output_umbrella/16/win16.colvars.traj 34 10
./output_umbrella/17/win17.colvars.traj 33 10
./output_umbrella/18/win18.colvars.traj 32 10
./output_umbrella/19/win19.colvars.traj 31 10
./output_umbrella/20/win20.colvars.traj 30 10
./output_umbrella/21/win21.colvars.traj 29 10
./output_umbrella/22/win22.colvars.traj 28 10
./output_umbrella/23/win23.colvars.traj 27 10
./output_umbrella/24/win24.colvars.traj 26 10
./output_umbrella/25/win25.colvars.traj 25 10
./output_umbrella/26/win26.colvars.traj 24 10
./output_umbrella/27/win27.colvars.traj 23 10
./output_umbrella/28/win28.colvars.traj 22 10
./output_umbrella/29/win29.colvars.traj 21 10
./output_umbrella/30/win30.colvars.traj 20 10
./output_umbrella/31/win31.colvars.traj 19 10
./output_umbrella/32/win32.colvars.traj 18 10
./output_umbrella/33/win33.colvars.traj 17 10
./output_umbrella/34/win34.colvars.traj 16 10
./output_umbrella/35/win35.colvars.traj 15 10
./output_umbrella/36/win36.colvars.traj 14 10
./output_umbrella/37/win37.colvars.traj 13 10
./output_umbrella/38/win38.colvars.traj 12 10
./output_umbrella/39/win39.colvars.traj 11 10
./output_umbrella/40/win40.colvars.traj 10 10
Please give me a suggestion or advise to solve this issue.
Thank you,
Jun

This archive was generated by hypermail 2.1.6 : Wed Dec 04 2019 - 23:20:28 CST