Re: ABF inFiles

From: Subramanian Vaitheeswaran (
Date: Thu Jun 21 2007 - 12:29:02 CDT

Hi Jerome,

Thank you for the prompt reply. I also use the InFiles entry to combine data from multiple trajectories in one final, short MD run.


---- Original message ----
>Date: Thu, 21 Jun 2007 12:57:35 -0400
>From: Jerome Henin <>
>Subject: Re: namd-l: ABF inFiles
>To: Subramanian Vaitheeswaran <>
>Hi Vaithee,
>Your first reaction is correct: since data read through niFiles gets included
>in the output file, in most cases one should only include the output from the
>previous run. We made it possible to read several files there only to make it
>easier for users to switch between several smaller "windows" and larger ones,
>merging the results of several previous runs.
>In addition to checking the number of samples in each bin, you can look in the
>output for this message:
>TCL: ABF> Loading previous force data from meoh01.dat ...
>TCL: ABF> 2454014 sampling points retrieved
>The number of samples there may be a bit smaller than the number of timesteps
>if the system has spent some time at the boundaries, where no samples are
>You are not the only person to have been confused about this, so I will think
>of updating the manual to make this point clearer.
>On Thursday 21 June 2007 12:03, Subramanian Vaitheeswaran wrote:
>> I have a question about combining the ABF output from multiple data sets
>> using the inFiles entry in the configuration file.
>> SET2 reads restart files from SET1 and ABF section of the configuration
>> file I have the following entry: abf inFiles {../SET1/abf.pmf}
>> So, SET2 will include data from SET1 in calculating the PMF.
>> SET3 then reads restart files from SET2. To combine the data from all
>> three sets, should the configuration file read: (1) abf inFiles
>> {../SET2/abf.pmf} or,
>> (2) abf inFiles {../SET1/abf.pmf ../SET2/abf.pmf} ?
>> I thought SET2/abf.pmf would include data from SET1, so it should be (1),
>> but looking at the number of samples in each bin, I am not sure any more.
>> Any help is much appreciated.
>> thanks,
>> Vaithee

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:44:52 CST