On Thu, Apr 16, 2020 at 9:27 AM Wolfgang Verestek <
[email protected]> wrote:

>
>
> Am Donnerstag, 16. April 2020 16:38:40 UTC+2 schrieb Christoph Junghans:
>>
>>
>>
>> On Thu, Apr 16, 2020 at 1:37 AM Wolfgang Verestek <[email protected]>
>> wrote:
>>
>>>
>>>
>>> Am Mittwoch, 15. April 2020 14:55:59 UTC+2 schrieb Christoph Junghans:
>>>>
>>>>
>>>>
>>>> On Wed, Apr 15, 2020 at 6:36 AM Wolfgang Verestek <
>>>> [email protected]> wrote:
>>>>
>>>>> There is indeed a slight difference. but as stated before, i think
>>>>> qualitatively they are matching.
>>>>>
>>>>> Second thing: I increased the cutoff to be more than half the box (9
>>>>> -> 15 A). This results in the behavior that we have seen before. Also the
>>>>> VMD
>>>>>
>>>> Ok, now I see, well the box size is 18.6 A, a cutoff of more than 9.3 A
>>>> isn't really physically meaningful and csg_stat should basically warn of
>>>> that.
>>>>
>>> ok, thanks. But could give me a hint why such a cutoff is physically not
>>> meaningful? I understand that such a small system does have bad statistics
>>> etc. but the particle is not self-interacting (although the molecule is,
>>> but it's also doing that with cutoff of 0.5* box length). Isn't that
>>> exactly what periodic boundaries should cover? I mean, extreme case here,
>>> but...?
>>>
>> Periodic boundaries follow minimum image convention, that means one only
>> counts minimum of all distances in all periodic images.
>> Imagine a 1D box of length L and put one particle at 0 and another at
>> L/2. Their distance will be L/2, but if you move it further away from 0, it
>> will come closer to the periodic image of the 1st particle sitting at L and
>> hence making the distance shorter again. In 3D there is another factor of
>> sqrt(3), but I think that is the reason why VMD had 0s in the rdf for
>> bigger distances.
>>
>> Christoph
>>
> Thanks for the explanation. I was not aware that every particle is
> considered just once.
>
> I might be wrong, but from my understanding two interactions in the
> periodic 1D case would be possible. E.g. for two particles "x" and "o" with
> periodic boundary "|": o would have one interactions with x for d<L/2 and a
> second one for 0.5L<d<L, the first at 2 units distance in positive/right
> direction and the second with at 7 units distance in negative/left
> direction. So if I understand you correct, only one image of the particle
> with the minimum distance is considered and all the other periodic images
> are neglected?
>
Yes, for the RDF only the first image is considered.

For force interactions it depends, Lennard-Jones usually only uses the
first image, but coulomb interactions, as they decay slower, sum over all
images (see Ewald summation).

Christoph


>
> Image   -1    |    0    |    1
> Dist     98765 432101234 56789
> Sys  ....o.x..|....o.x..|....o.x..
>
> Wolfgang
>
>>
>>>
>>>>
>>>> Christoph
>>>>
>>>>>
>>>>>
>>>>> Am Mittwoch, 15. April 2020 14:13:30 UTC+2 schrieb Christoph Junghans:
>>>>>>
>>>>>>
>>>>>>
>>>>>> On Wed, Apr 15, 2020 at 6:05 AM Wolfgang Verestek <
>>>>>> [email protected]> wrote:
>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Am Mittwoch, 15. April 2020 14:00:32 UTC+2 schrieb Christoph
>>>>>>> Junghans:
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> On Wed, Apr 15, 2020 at 05:56 Wolfgang Verestek <
>>>>>>>> [email protected]> wrote:
>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> Am Mittwoch, 15. April 2020 13:52:01 UTC+2 schrieb Wolfgang
>>>>>>>>> Verestek:
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>> Am Dienstag, 14. April 2020 19:21:07 UTC+2 schrieb Christoph
>>>>>>>>>> Junghans:
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>> On Tue, Apr 14, 2020 at 8:40 AM Wolfgang Verestek wrote:
>>>>>>>>>>>
>>>>>>>>>>>> Dear Christoph,
>>>>>>>>>>>>
>>>>>>>>>>>> it seems to me, that the bonds in topology.xml do not influence
>>>>>>>>>>>> the result of csg_stat. On contrary, i got the same results 
>>>>>>>>>>>> regardless of
>>>>>>>>>>>> bonds being in topology.xml or not. The same is true for for 
>>>>>>>>>>>> wrapped or
>>>>>>>>>>>> unwrapped coordinates. The only thing that did change the result 
>>>>>>>>>>>> was
>>>>>>>>>>>> changing the cell size in the conf.gro by factor of 10 in each 
>>>>>>>>>>>> dimenstion
>>>>>>>>>>>> which looks like the nonperiodic results from Ovito (but made it 
>>>>>>>>>>>> necessary
>>>>>>>>>>>> to scale by a factor of 1000 to be in the same range of the plot).
>>>>>>>>>>>> Furthermore, i confirmed the results of Ovito by VMD. The 
>>>>>>>>>>>> nonperiodic
>>>>>>>>>>>> calculation in VMD oversitmates the first peak somewhat, but 
>>>>>>>>>>>> otherwise the
>>>>>>>>>>>> results are in good agreement. The results don't match perfectly, 
>>>>>>>>>>>> but as i
>>>>>>>>>>>> don't know how exactly VMD or Ovito compute the RDF i guess the 
>>>>>>>>>>>> match is
>>>>>>>>>>>> good enough and the drop in the periodic VMD plot should be due to 
>>>>>>>>>>>> some
>>>>>>>>>>>> other reason (maybe the g(r) plugin is limited to 3/4 cell lengt 
>>>>>>>>>>>> or sth.).
>>>>>>>>>>>> regarding scg_stat it still seems to me that the tail is
>>>>>>>>>>>> smoothed or periodic images in different directions are omitted. 
>>>>>>>>>>>> Could that
>>>>>>>>>>>> be? I can provide the files f necessary.
>>>>>>>>>>>>
>>>>>>>>>>> Ok now I am getting really confused, I just did a quick
>>>>>>>>>>> comparison of gmx rdf and csg_stat and the results are right on top 
>>>>>>>>>>> of each
>>>>>>>>>>> other.
>>>>>>>>>>>
>>>>>>>>>>> Could you do me a favor and try to follow:
>>>>>>>>>>> ```
>>>>>>>>>>> apt-get install gromacs
>>>>>>>>>>> cp /usr/share/gromacs/top/spc216.gro .
>>>>>>>>>>> wget
>>>>>>>>>>> https://raw.githubusercontent.com/votca/csg/master/src/tools/references/spce/sources/grompp.mdp
>>>>>>>>>>> wget
>>>>>>>>>>> https://raw.githubusercontent.com/votca/csg/master/src/tools/references/spce/sources/topol.top
>>>>>>>>>>> gmx grompp -c spc216.gro -maxwarn 1
>>>>>>>>>>> echo q | gmx make_ndx -f spc216.gro
>>>>>>>>>>> echo -e "1\n1" | gmx rdf -f spc216.gro -s topol.tpr -n index.ndx
>>>>>>>>>>> -selrpos mol_com -seltype mol_com -bin 0.02
>>>>>>>>>>> wget
>>>>>>>>>>> https://raw.githubusercontent.com/votca/csg/master/src/tools/references/spce/mapping.xml
>>>>>>>>>>> wget
>>>>>>>>>>> https://raw.githubusercontent.com/votca/csg/master/src/tools/references/spce/settings_rdf.xml
>>>>>>>>>>> csg_stat --options settings_rdf.xml --top  topol.tpr --trj
>>>>>>>>>>> spc216.gro --cg mapping.xml
>>>>>>>>>>> csg_map --top  topol.tpr --trj spc216.gro --cg mapping.xml --out
>>>>>>>>>>> cg.dump
>>>>>>>>>>> ```
>>>>>>>>>>> and then calculate the rdf from cg.dump with ovito and vmd and
>>>>>>>>>>> compare them with rdf.xvg and CG-CG.dist.new.
>>>>>>>>>>>
>>>>>>>>>>> Christoph
>>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>> Ovito and VMD give nearly the same results (green X / blue
>>>>>>>>>> square). I'm not sure how those two really compute the RDF. For 
>>>>>>>>>> Ovito I can
>>>>>>>>>> set the cutoff and the number of bins (9.0, 45) and i think the bin 
>>>>>>>>>> center
>>>>>>>>>> is on the x coordinates. In VMD i can set the cutoff and the delta 
>>>>>>>>>> (9,
>>>>>>>>>> 0.2). As said, i'm not sure how the bins are decided inside those two
>>>>>>>>>> progs, so this might lead to small offsets.
>>>>>>>>>> cg_stat on the other hand is a little bit off compared to ovito
>>>>>>>>>> and vmd. But this could be imho due to the binning as mentioned 
>>>>>>>>>> before.
>>>>>>>>>> qualitatively i think it's pretty much the same.
>>>>>>>>>>
>>>>>>>>> You can pick the bins size for csg_stat in settings_rdf.xml and
>>>>>>>> for gmx on command line so you can make it same as vmd for the 
>>>>>>>> comparison.
>>>>>>>>
>>>>>>>> Christoph
>>>>>>>>
>>>>>>>
>>>>>>> bin size is the same. 0.2 in vmd, 0.02 in settings_rdf.xml.
>>>>>>> I forgot the plot in the last post
>>>>>>>
>>>>>> Hmm, but it seems the dots are not at the same r values, or am I
>>>>>> seeing this wrong?
>>>>>>
>>>>>> Christoph
>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>>>> - regards
>>>>>>>>>> Wolfgang
>>>>>>>>>>
>>>>>>>>> forgot the plot
>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>> Am Mittwoch, 8. April 2020 18:17:20 UTC+2 schrieb Christoph
>>>>>>>>>>>> Junghans:
>>>>>>>>>>>>>
>>>>>>>>>>>>>
>>>>>>>>>>>>>
>>>>>>>>>>>>> On Wed, Apr 8, 2020 at 5:01 AM Wolfgang Verestek <
>>>>>>>>>>>>> [email protected]> wrote:
>>>>>>>>>>>>>
>>>>>>>>>>>>>>
>>>>>>>>>>>>>>
>>>>>>>>>>>>>> Am Mittwoch, 8. April 2020 02:07:39 UTC+2 schrieb Christoph
>>>>>>>>>>>>>> Junghans:
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>> On Tue, Apr 7, 2020 at 9:59 AM Wolfgang Verestek <
>>>>>>>>>>>>>>> [email protected]> wrote:
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>> Dear Christoph,
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>> thanks for your fast response. i followed you suggestion,
>>>>>>>>>>>>>>>> renamed the files and added the box tag to the topology file 
>>>>>>>>>>>>>>>> (files
>>>>>>>>>>>>>>>> attached). But the result stays the same. So, as you said that 
>>>>>>>>>>>>>>>> gromacs
>>>>>>>>>>>>>>>> format is better supported i changed my output to gro87 
>>>>>>>>>>>>>>>> format. No errors,
>>>>>>>>>>>>>>>> but the discrepancy remains, regardless if i'm using wrapped 
>>>>>>>>>>>>>>>> or unwrapped
>>>>>>>>>>>>>>>> coords in the conf.gro file. See below. Are there any other 
>>>>>>>>>>>>>>>> things i could
>>>>>>>>>>>>>>>> change? Another format?
>>>>>>>>>>>>>>>> Is it possible to set the boundary conditions somewhere
>>>>>>>>>>>>>>>> specifically? Or could it be that periodic images of the 
>>>>>>>>>>>>>>>> specific atoms are
>>>>>>>>>>>>>>>> ignored (the simulation box is kind of small and close by 
>>>>>>>>>>>>>>>> region seems ok)?
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>> Sorry, I didn't have time to look at your files yet, but
>>>>>>>>>>>>>>> does OVito handle exclusion in the same way as VOTCA?
>>>>>>>>>>>>>>> Have a look at the output of "csg_dump --top topology.xml
>>>>>>>>>>>>>>> --excl"
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>
>>>>>>>>>>>>>> How should that look like? I'm not sure if i understand that
>>>>>>>>>>>>>> format correctly, but it seems correct to me. in my data file 
>>>>>>>>>>>>>> Atom with id
>>>>>>>>>>>>>> 1 is connectod to atoms 2 3 4 and 36. I assume this is non 
>>>>>>>>>>>>>> repetitive so id
>>>>>>>>>>>>>> 4 does not show up again for atom id 4 and so on. I did not 
>>>>>>>>>>>>>> check the whole
>>>>>>>>>>>>>> list, but the first 10 lines are correct.
>>>>>>>>>>>>>>
>>>>>>>>>>>>>> I have 164 beads in 1 molecules
>>>>>>>>>>>>>>
>>>>>>>>>>>>>> List of exclusions:
>>>>>>>>>>>>>> 1 2 3 4 36
>>>>>>>>>>>>>>
>>>>>>>>>>>>> This means the atom #1 is excluded with atom #2, #3,#4 and #36
>>>>>>>>>>>>> and hence contributions from these pairs won't go into the RDF.
>>>>>>>>>>>>> VOTCA automatically created exclusion (in same way as Gromacs
>>>>>>>>>>>>> does) for bonds, angles and dihedrals.
>>>>>>>>>>>>>
>>>>>>>>>>>>> Christoph
>>>>>>>>>>>>>
>>>>>>>>>>>>>
>>>>>>>>>>>>>> 4 5 6 7
>>>>>>>>>>>>>> 7 8 9 10
>>>>>>>>>>>>>> 10 11 12
>>>>>>>>>>>>>> 12 13 14
>>>>>>>>>>>>>> 14 15 16 17
>>>>>>>>>>>>>> 17 18 19 39
>>>>>>>>>>>>>> 20 21 22 23 77
>>>>>>>>>>>>>> ...
>>>>>>>>>>>>>>
>>>>>>>>>>>>>>
>>>>>>>>>>>>>>> Secondly you could use "csg_map --top topology --no-map
>>>>>>>>>>>>>>> --trj lammps.dump --out out.pdb", to check if the coordinates 
>>>>>>>>>>>>>>> are read
>>>>>>>>>>>>>>> correctly.
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>> apart from the bigger cell for the pdb (where this
>>>>>>>>>>>>>> information is not included) the positions look ok. (lammps file 
>>>>>>>>>>>>>> in "black
>>>>>>>>>>>>>> cores", pdb in transparent green).
>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>> Christoph
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>> Is there a way to visualize the xml or gro files? I know
>>>>>>>>>>>>>>>> vmd, but that segfaults when reading the gro file.
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>> Ultimately the whole thing should go into force matching.
>>>>>>>>>>>>>>>> But i guess the results will be off if already the RDF is 
>>>>>>>>>>>>>>>> wrong?
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>> Is it really necessary to have one file per molecule type
>>>>>>>>>>>>>>>> and what distinguishes different molecule types? If i have 
>>>>>>>>>>>>>>>> several PA6
>>>>>>>>>>>>>>>> molecules with different length/count of monomers is that 
>>>>>>>>>>>>>>>> already different
>>>>>>>>>>>>>>>> molecules?
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>> Thanks and regards
>>>>>>>>>>>>>>>> Wolfgang
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>> --
>>>>>>>>>>>>>>>> You received this message because you are subscribed to the
>>>>>>>>>>>>>>>> Google Groups "votca" group.
>>>>>>>>>>>>>>>> To unsubscribe from this group and stop receiving emails
>>>>>>>>>>>>>>>> from it, send an email to [email protected].
>>>>>>>>>>>>>>>> To view this discussion on the web visit
>>>>>>>>>>>>>>>> https://groups.google.com/d/msgid/votca/b4b7484f-2663-4fe6-a082-ad3c2ca1ef20%40googlegroups.com
>>>>>>>>>>>>>>>> <https://groups.google.com/d/msgid/votca/b4b7484f-2663-4fe6-a082-ad3c2ca1ef20%40googlegroups.com?utm_medium=email&utm_source=footer>
>>>>>>>>>>>>>>>> .
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>> --
>>>>>>>>>>>>>>> Christoph Junghans
>>>>>>>>>>>>>>> Web: http://www.compphys.de
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>> --
>>>>>>>>>>>>>> You received this message because you are subscribed to the
>>>>>>>>>>>>>> Google Groups "votca" group.
>>>>>>>>>>>>>> To unsubscribe from this group and stop receiving emails from
>>>>>>>>>>>>>> it, send an email to [email protected].
>>>>>>>>>>>>>> To view this discussion on the web visit
>>>>>>>>>>>>>> https://groups.google.com/d/msgid/votca/3cd73547-59da-4de8-894d-0345302ebeca%40googlegroups.com
>>>>>>>>>>>>>> <https://groups.google.com/d/msgid/votca/3cd73547-59da-4de8-894d-0345302ebeca%40googlegroups.com?utm_medium=email&utm_source=footer>
>>>>>>>>>>>>>> .
>>>>>>>>>>>>>>
>>>>>>>>>>>>>
>>>>>>>>>>>>>
>>>>>>>>>>>>> --
>>>>>>>>>>>>> Christoph Junghans
>>>>>>>>>>>>> Web: http://www.compphys.de
>>>>>>>>>>>>>
>>>>>>>>>>>> --
>>>>>>>>>>>> Join us on Slack: https://join.slack.com/t/votca/signup
>>>>>>>>>>>> ---
>>>>>>>>>>>> You received this message because you are subscribed to the
>>>>>>>>>>>> Google Groups "votca" group.
>>>>>>>>>>>> To unsubscribe from this group and stop receiving emails from
>>>>>>>>>>>> it, send an email to [email protected].
>>>>>>>>>>>> To view this discussion on the web visit
>>>>>>>>>>>> https://groups.google.com/d/msgid/votca/9c1da74b-2bca-4aa5-b1a1-f26123328000%40googlegroups.com
>>>>>>>>>>>> <https://groups.google.com/d/msgid/votca/9c1da74b-2bca-4aa5-b1a1-f26123328000%40googlegroups.com?utm_medium=email&utm_source=footer>
>>>>>>>>>>>> .
>>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>> --
>>>>>>>>>>> Christoph Junghans
>>>>>>>>>>> Web: http://www.compphys.de
>>>>>>>>>>>
>>>>>>>>>> --
>>>>>>>>> Join us on Slack: https://join.slack.com/t/votca/signup
>>>>>>>>> ---
>>>>>>>>> You received this message because you are subscribed to the Google
>>>>>>>>> Groups "votca" group.
>>>>>>>>> To unsubscribe from this group and stop receiving emails from it,
>>>>>>>>> send an email to [email protected].
>>>>>>>>> To view this discussion on the web visit
>>>>>>>>> https://groups.google.com/d/msgid/votca/12ef52f9-854d-4404-95e5-54437779ae7a%40googlegroups.com
>>>>>>>>> <https://groups.google.com/d/msgid/votca/12ef52f9-854d-4404-95e5-54437779ae7a%40googlegroups.com?utm_medium=email&utm_source=footer>
>>>>>>>>> .
>>>>>>>>>
>>>>>>>> --
>>>>>>>> Christoph Junghans
>>>>>>>> Web: http://www.compphys.de
>>>>>>>>
>>>>>>> --
>>>>>>> Join us on Slack: https://join.slack.com/t/votca/signup
>>>>>>> ---
>>>>>>> You received this message because you are subscribed to the Google
>>>>>>> Groups "votca" group.
>>>>>>> To unsubscribe from this group and stop receiving emails from it,
>>>>>>> send an email to [email protected].
>>>>>>> To view this discussion on the web visit
>>>>>>> https://groups.google.com/d/msgid/votca/56cd974d-7a71-470f-b895-d95163c22596%40googlegroups.com
>>>>>>> <https://groups.google.com/d/msgid/votca/56cd974d-7a71-470f-b895-d95163c22596%40googlegroups.com?utm_medium=email&utm_source=footer>
>>>>>>> .
>>>>>>>
>>>>>>
>>>>>>
>>>>>> --
>>>>>> Christoph Junghans
>>>>>> Web: http://www.compphys.de
>>>>>>
>>>>> --
>>>>> Join us on Slack: https://join.slack.com/t/votca/signup
>>>>> ---
>>>>> You received this message because you are subscribed to the Google
>>>>> Groups "votca" group.
>>>>> To unsubscribe from this group and stop receiving emails from it, send
>>>>> an email to [email protected].
>>>>> To view this discussion on the web visit
>>>>> https://groups.google.com/d/msgid/votca/6a39277e-ab3a-4e14-9a8b-831a1ddb24d1%40googlegroups.com
>>>>> <https://groups.google.com/d/msgid/votca/6a39277e-ab3a-4e14-9a8b-831a1ddb24d1%40googlegroups.com?utm_medium=email&utm_source=footer>
>>>>> .
>>>>>
>>>>
>>>>
>>>> --
>>>> Christoph Junghans
>>>> Web: http://www.compphys.de
>>>>
>>>> --
>>> Join us on Slack: https://join.slack.com/t/votca/signup
>>> ---
>>> You received this message because you are subscribed to the Google
>>> Groups "votca" group.
>>> To unsubscribe from this group and stop receiving emails from it, send
>>> an email to [email protected].
>>> To view this discussion on the web visit
>>> https://groups.google.com/d/msgid/votca/6152ba41-02d8-4400-ade4-1f1b2e27f13d%40googlegroups.com
>>> <https://groups.google.com/d/msgid/votca/6152ba41-02d8-4400-ade4-1f1b2e27f13d%40googlegroups.com?utm_medium=email&utm_source=footer>
>>> .
>>>
>>
>>
>> --
>> Christoph Junghans
>> Web: http://www.compphys.de
>>
> --
> Join us on Slack: https://join.slack.com/t/votca/signup
> ---
> You received this message because you are subscribed to the Google Groups
> "votca" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to [email protected].
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/votca/11968a7e-fb3e-443d-8af5-a2773e241335%40googlegroups.com
> <https://groups.google.com/d/msgid/votca/11968a7e-fb3e-443d-8af5-a2773e241335%40googlegroups.com?utm_medium=email&utm_source=footer>
> .
>


-- 
Christoph Junghans
Web: http://www.compphys.de

-- 
Join us on Slack: https://join.slack.com/t/votca/signup
--- 
You received this message because you are subscribed to the Google Groups 
"votca" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To view this discussion on the web visit 
https://groups.google.com/d/msgid/votca/CAHG27e6Qb3qYCjJC7g9pPjecS9DV4ri8QCaToT3R_7MY%2BC0sLA%40mail.gmail.com.

Reply via email to