[gmx-users] .dat file from gmx densmap doestnottmatchtprescribedtsize

Michele Pellegrino micpel at kth.se
Thu Jan 23 11:02:31 CET 2020


Ok, I believe I have found the nature of this 'quirk' looking at the source code (as Berk suggested me to do).
I was assuming a format, but GROMACS is producing another: the .dat file in the output contains the values PLUS the coordinates for each bin, that is the first row and the first colum are not denisty values, but coordinates; hence the output matrix is indeed (n1+1)x(n2+1), the actual values being in grid[1:n1-1,2:n2-2] (0-start index convention).
So apparently is not a bug in the code, but just a misunderstanding of me as user; I double checked the documentation (both for 2019.04 and for the new 2020 release) and I did not find any reference to the format of the output file: could it be a good idea to incude some indication in that regard?

Cheers,
Michele
________________________________________
From: gromacs.org_gmx-users-bounces at maillist.sys.kth.se <gromacs.org_gmx-users-bounces at maillist.sys.kth.se> on behalf of Michele Pellegrino <micpel at kth.se>
Sent: 22 January 2020 22:58
To: gmx-users at gromacs.org
Subject: Re: [gmx-users] .dat file from gmx densmap doestnottmatchtprescribedtsize

I tried just now: it did not work: 1728 values expected, 1645 values in the .dat file.
I managed to make it work by defining the number of values in each direction as (in Python):
Nx = np.sqrt(Nda*Lx/Lz)
Nz = Nx*Lz/Lx
Nx = int( np.round(Nx) )
Nz = int( np.round(Nz) )
being Nda the number of values in the .dat file and Lx (resp. Lz) the lenght of the simulation box in the x (z) direction (np stands for numpy).
However, I believe this manipulation should not be necessary.

Cheers,
Michele
________________________________________
From: gromacs.org_gmx-users-bounces at maillist.sys.kth.se <gromacs.org_gmx-users-bounces at maillist.sys.kth.se> on behalf of David van der Spoel <spoel at xray.bmc.uu.se>
Sent: 22 January 2020 22:07
To: gmx-users at gromacs.org
Subject: Re: [gmx-users] .dat file from gmx densmap doestnottmatchtprescribedtsize

Den 2020-01-22 kl. 18:44, skrev Michele Pellegrino:
> Ok, thank you.
> I am counting the elements assuming they are in an array organized in a row major fashion.
have you tried supplying -n1 47 -n2 35 to be able to check manually?

(by the way a resolution of 0.1 Angstrom is unlikely to give you a
smooth surface)
>
> Scarica Outlook per Android<https://aka.ms/ghei36>
> ________________________________
> From: gromacs.org_gmx-users-bounces at maillist.sys.kth.se <gromacs.org_gmx-users-bounces at maillist.sys.kth.se> on behalf of Berk Hess <gmx3 at hotmail.com>
> Sent: Wednesday, January 22, 2020 6:34:47 PM
> To: gmx-users at gromacs.org <gmx-users at gromacs.org>
> Subject: Re: [gmx-users] .dat file from gmx densmap does not match prescribed size
>
> A grid element
>
> I looked at the code and it directly uses n1 and n2 you provide. So I thought that it might be that the code reading the file has a limit on the number of characters on a line.
>
> /Berk
>
> ________________________________
> From: gromacs.org_gmx-users-bounces at maillist.sys.kth.se <gromacs.org_gmx-users-bounces at maillist.sys.kth.se> on behalf of Michele Pellegrino <micpel at kth.se>
> Sent: Wednesday, January 22, 2020 5:07 PM
> To: gmx-users at gromacs.org <gmx-users at gromacs.org>
> Subject: Re: [gmx-users] .dat file from gmx densmap does not match prescribed size
>
>
> What do you mean by 'element'?
>
> Anyway, part of the issue was in how I was reading the file; with a little modification of the script I reduced the mismatch, but it still persists.
> Could it be because GROMACS forces a grid with same meshwith in x and (in my case) z direction, so that the used should impose either -n1 OR -n2 (but not both)? Or maybe the issue is just in the format of the .dat output file?
>
> By the way, the simulation box is 47.72970  x 5.51140 x 35.37240 nanometers (if that may help understanding the issue).
>
> Thanks,
> Michele
>
> ________________________________________
> From: gromacs.org_gmx-users-bounces at maillist.sys.kth.se <gromacs.org_gmx-users-bounces at maillist.sys.kth.se> on behalf of Berk Hess <gmx3 at hotmail.com>
> Sent: 22 January 2020 16:49
> To: gmx-users at gromacs.org
> Subject: Re: [gmx-users] .dat file from gmx densmap does not match      prescribed      size
>
> How are you counting the elements? Lines with 3531 elements are very long, so they can easily exceed some limit.
>
> /Berk
>
> ________________________________
> From: gromacs.org_gmx-users-bounces at maillist.sys.kth.se <gromacs.org_gmx-users-bounces at maillist.sys.kth.se> on behalf of Michele Pellegrino <micpel at kth.se>
> Sent: Tuesday, January 21, 2020 5:16 PM
> To: gromacs.org_gmx-users at maillist.sys.kth.se <gromacs.org_gmx-users at maillist.sys.kth.se>
> Subject: [gmx-users] .dat file from gmx densmap does not match prescribed size
>
> Hello,
>
>
> I am struggling a little bit with gmx densmap. I am trying to obtain a matrix with density values from the .dat file generated by the program; however the number of data in the output file does not match what I am expecting.
>
> I tried to prescribe both the number of bins in each direction (-n1 and -n2 options) and the grid size (-bin option), not at the same, of course:
>
>
> gmx densmap -f traj.trr -s system.tpr -od densmap.dat -b 800.0 -e 1200.0 -aver y -n1 4772 -n2 3531
>
>
> OR (equivalently):
>
>
> gmx densmap -f traj.trr -s system.tpr -od densmap.dat -b 800.0 -e 1200.0 -aver y -bin 0.01
>
>
> In both cases I get a dimension mismatch (I expect 16849932 values but I get 4359553).
>
> I am using GROMACS 2019.4?.
>
>
> What am I doing wrong?
>
>
> Thanks for your help.
>
>
> Cheers,
>
> Michele
> --
> Gromacs Users mailing list
>
> * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!
>
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>
> * For (un)subscribe requests visit
> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-request at gromacs.org.
> --
> Gromacs Users mailing list
>
> * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!
>
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>
> * For (un)subscribe requests visit
> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-request at gromacs.org.
> --
> Gromacs Users mailing list
>
> * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!
>
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>
> * For (un)subscribe requests visit
> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-request at gromacs.org.
> --
> Gromacs Users mailing list
>
> * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!
>
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>
> * For (un)subscribe requests visit
> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-request at gromacs.org.
>


--
David van der Spoel, Ph.D., Professor of Biology
Head of Department, Cell & Molecular Biology, Uppsala University.
Box 596, SE-75124 Uppsala, Sweden. Phone: +46184714205.
http://www.icm.uu.se
--
Gromacs Users mailing list

* Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!

* Can't post? Read http://www.gromacs.org/Support/Mailing_Lists

* For (un)subscribe requests visit
https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-request at gromacs.org.
--
Gromacs Users mailing list

* Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!

* Can't post? Read http://www.gromacs.org/Support/Mailing_Lists

* For (un)subscribe requests visit
https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-request at gromacs.org.


More information about the gromacs.org_gmx-users mailing list