[gmx-users] g_bar bug with temperature

[Mark Abraham]

On 6/03/2012 6:58 PM, Tom Kirchner wrote:
> Dear all,
>
> I made a free energy calculation and want to evaluate my data with
> g_bar. But when I am using g_bar on my .xvg files I get an error of the
> form:
>
>      Temperature in file md_0.35.xvg different from earlier files or 
> setting
>
> All simulations were performed at temperature 298.15 K, this number is
> also written in the .xvg files and the same for all of them. An older
> setup worked fine a the temperature of 300 K. After I changed the
> temperature of my new files also to 300 K, g_bar now works.
>
> Is it possible, that g_bar can not handle real numbers?

Unlikely. You can perhaps probe that by using 300.0 and 300.1.

When suggesting the existence of a bug, people need to be explicit about 
the conditions that might replicate it - so GROMACS version and actual 
command lines and (here) copies of headers of .xvg files are absolutely 
necessary. To get traction, you need to overcome the normal assumption 
of the developer that they've coded it right and you're using it wrongly 
:-) Information is the key to achieving that.

Mark