[CP2K-user] [CP2K:20199] Re: unable to analyse the metadynamics output file of the cp2k

[David M. Sherman]

You're right: the tutorials don't see to explain this bit at all...

You need to use the "graph" program which comes with cp2k. When you do a 
metadynamics run, the resulting *.restart file will contain a record of all 
of the HILLS  added during the metadynamics simulation. Note that if you do 
a multiple-walker run, the *.restart file in any one of the walker 
directories will also have all the hills from all the other walkers. Let's 
assume you did a metadynamics simulation with 2 COLVARS and this generated 
a restart file named "mysim.restart".  To turn that record of HILLS in the 
restart file into a one-dimensional free-energy surface with respect to the 
first COLVAR, you run the program graph.psmp using the following syntax:

graph.psmp -cp2k -ndim 2 -ndw 1 -file mysim.restart

If you want  a one-dimensional free energy surface with respect to the 
second COLVAR, you run

graph.psmp -cp2k -ndim 2 -ndw 2  -file mysim.restart

(Note: BE VERY CAREFUL HERE!  Even though you only want a one dimensional 
fes, the argument to -ndim MUST be the actual number of COLVARS that were 
used in the metadynamics run.  The argument to -ndw is which COLVAR(S) you 
want to generate the free-energy surface with respect to.  If you 
accidently use -ndim 1, you will get what looks like a perfectly sensible 
fes, but it will be completely erroneous).

Running the graph program will generate a file called fes.dat that you can 
plot with gnuplot to see the free energy surface.  Here's an example of a 
gnuplot script to
generate the plot (note: I'm converting my Cl coordination number COLVAR 
into a sensible number by multiplying it by the number of Cl atoms in the 
cell  and converting the free energy from AU to kJ/mole).

set xlabel "Cl CN"
#set terminal svg
set xrange[-1:4]
set yrange[-230:-120]
plot "fes.dat" using ($1*4):($2*2625.5) with lines
pause -1 "Hit any key to continue"

Now,  if you want a 2-dimensional plot of the the free energy surface 
(i.e., over both
COLVARS) then you run:

graph.psmp -cp2k -ndim 2 -ndw 1 2 -file mysim.restart

and plot that using gnuplot.  Here's a gnuplot input script to generate a 
contour map:

set view map
unset surface
set contour
set cntrparam levels 30
set dgrid3d 100,100,1
set xlabel "Cl CN"
set ylabel "O CN"
set style line 1 lc rgb "black" lw 1
set style increment user
#set terminal svg
splot [-1:4] [-1:5] "fes.dat" using ($1*4.):($2*55):($3*2625.5) with lines
pause -1 "Hit any key to continue"

Have fun!

On Tuesday, May 7, 2024 at 10:33:31 PM UTC+1 Karl Irikura wrote:

> I have the same question.  That tutorial does not explain how to extract 
> free energy information from the output files.  
>
> Thanks for any tips,
> Karl
>
> On Thursday, January 11, 2024 at 7:52:20 AM UTC-5 Marcella Iannuzzi wrote:
>
>> Please refer to some tutorial like
>> https://www.cp2k.org/exercises:2015_cecam_tutorial:mtd1
>>
>> Regards
>> Marcella
>>
>>
>> On Wednesday, January 10, 2024 at 1:02:01 PM UTC+1 cy22res... at iith.ac.in 
>> wrote:
>>
>>> Hey Good afternoon everyone
>>> Can anybody known how to analysis the cp2k metadynamics file? Please 
>>> help if possible
>>> Thank you
>>> Deewan
>>>
>>

-- 
You received this message because you are subscribed to the Google Groups "cp2k" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cp2k+unsubscribe at googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/b0a8a14b-d8ae-4359-8573-e84755bbd1edn%40googlegroups.com.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20240509/bc8467d1/attachment.htm>