[CP2K-user] [CP2K:19066] Unexpected 0D wavelet errors at large vacuum

[Léon Luntadila Lufungula]

Hi Jurg,

I guess it's a bug in the 2023.1 version then, because I also just tested 
the most recent pre-compiled binary 
(https://github.com/cp2k/cp2k/releases/download/v2023.1/cp2k-2023.1-Linux-gnu-x86_64.ssmp) 
and it still gives the same errors:

[image: pre-compiled binary.png]

Is a new version (2023.2) coming anytime soon perhaps? I tried compiling 
the trunk earlier this week after you fixed the CIF parser, both through 
cloning the Master from Git and using the tarball, but got errors during 
compilation... I'm going on vacation tomorrow but after I return I'll give 
it another try or if a newer version has released by then I'll try that and 
convince the HPC centre to install it.

Best,
Léon

On Wednesday, 5 July 2023 at 15:03:44 UTC+2 Jürg Hutter wrote:

> Hi
>
> it works for me (CP2K Trunk version)
>
> Precond_solver : DEFAULT
> stepsize : 0.08000000 energy_gap : 0.08000000
> eps_taylor : 0.10000E-15 max_taylor : 4
> ----------------------------------- OT 
> ---------------------------------------
>
> Step Update method Time Convergence Total energy Change
>
> ------------------------------------------------------------------------------
> 1 OT DIIS 0.80E-01 7.1 0.10459074 -83.0126092838 -8.30E+01
> 2 OT DIIS 0.80E-01 7.0 0.07934045 -85.4719301173 -2.46E+00
> 3 OT DIIS 0.80E-01 7.3 0.04237559 -86.5088026249 -1.04E+00
> 4 OT DIIS 0.80E-01 7.3 0.02451275 -87.1588661747 -6.50E-01
> 5 OT DIIS 0.80E-01 7.3 0.01463005 -87.2604337873 -1.02E-01
> 6 OT DIIS 0.80E-01 7.3 0.00701148 -87.3511713376 -9.07E-02
> 7 OT DIIS 0.80E-01 7.3 0.00477888 -87.3692546957 -1.81E-02
> 8 OT DIIS 0.80E-01 7.3 0.00289219 -87.3773030540 -8.05E-03
>
> This is running your input (removed the tabs and changed the file path).
>
> regards
> JH
>
> ________________________________________
> From: cp... at googlegroups.com <cp... at googlegroups.com> on behalf of Léon 
> Luntadila Lufungula <Leon.luntad... at uantwerpen.be>
> Sent: Wednesday, July 5, 2023 2:57 PM
> To: cp2k
> Subject: Re: [CP2K:19064] Unexpected 0D wavelet errors at large vacuum
>
> Just now, I also let CP2K print the atomic coordinates and the cell 
> information so that I could visualize it in VMD to see if perhaps something 
> goes wrong with reading in the structure. However, as you can see below, 
> there is nothing wrong with the structure and there is ample vacuum space 
> for the wavelet solver to work without throwing errors. I really don't 
> understand how these errors occur for a structure which is essentially 
> "over-qualified" for the wavelet method with 10 Å vacuum on all sides... Is 
> this perhaps a bug in the code behind the wavelet solver?
> [3mppa-vac-10.png]
>
> Best,
> Léon
> On Wednesday, 5 July 2023 at 14:07:38 UTC+2 Léon Luntadila Lufungula wrote:
> Hi Jurg,
>
> I still get the same errors when I alter the input as suggested, see the 
> files in attachment. In which version did you do the calculation? My 
> version of CP2K is 2023.1 and it seems to still throw errors even when 
> altering the file as suggested.
>
> Kind regards,
> Léon
>
> On Wednesday, 5 July 2023 at 13:57:53 UTC+2 Jürg Hutter wrote:
> Hi
>
> works for me. Only change was to set the CELL in the input file:
>
> &CELL
> ABC [angstrom] 25.000 25.000 25.000
> PERIODIC NONE
> &END CELL
>
> regards
> JH
>
> ________________________________________
> From: cp... at googlegroups.com <cp... at googlegroups.com> on behalf of Léon 
> Luntadila Lufungula <Leon.luntad... at uantwerpen.be>
> Sent: Wednesday, July 5, 2023 1:41 PM
> To: cp2k
> Subject: Re: [CP2K:19060] Unexpected 0D wavelet errors at large vacuum
>
> Hi Jurg,
>
> Thanks for the tip, but I already tried this and it didn't work. I have 
> tried once more and the results are in attachment... Any other suggestions?
>
> Best,
> Léon
>
> On Wednesday, 5 July 2023 at 13:03:51 UTC+2 Jürg Hutter wrote:
> Hi
>
> add center coordinates to your Topolgy section
>
> &TOPOLOGY
> COORD_FILE_NAME mol.cif
> COORD_FILE_FORMAT cif
> &CENTER_COORDINATES
> &END
> &END TOPOLOGY
>
> regards
> JH
>
> ________________________________________
> From: cp... at googlegroups.com <cp... at googlegroups.com> on behalf of Léon 
> Luntadila Lufungula <Leon.luntad... at uantwerpen.be>
> Sent: Wednesday, July 5, 2023 12:03 PM
> To: cp2k
> Subject: [CP2K:19057] Unexpected 0D wavelet errors at large vacuum
>
> Hi all,
>
> I have a quick question regarding the 0D wavelet solver. I tried using it 
> to calculate the energy of an isolated molecule and noticed some weird 
> behaviour regarding the vacuum padding. At 2.5 Å vacuum around the 
> molecule, I got an error saying "density non-zero on the edges of the unit 
> cell: wrong results in WAVELET solver", so I increased the vacuum to 5 Å 
> and this suppressed the error message. I wanted to see the effect of the 
> vacuum on the energy so I also increased the vacuum to compare the 
> energies. At 7.5 Å it all worked as expected, but then at 10 Å I again get 
> errors stating that the density is non-zero at the edges, even when I 
> include CENTER_COORDINATES. Can someone explain why this is happening? It 
> seems very counterintuitive as it is exactly the same molecule and the only 
> difference is that the vacuum padding around the molecule has increased...
>
> The outputs for 5, 7.5 and 10 Å are in attachment. The one with 2.5 Å, I 
> seem to have deleted already as it resulting in the errors stated above.
>
> Best,
> Léon
>
> --
> 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+uns... at googlegroups.com<mailto:cp2k+uns... at googlegroups.com
> >.
> To view this discussion on the web visit 
> https://groups.google.com/d/msgid/cp2k/589fb184-702d-485e-b911-5a75da8dae66n%40googlegroups.com
> <
> https://groups.google.com/d/msgid/cp2k/589fb184-702d-485e-b911-5a75da8dae66n%40googlegroups.com?utm_medium=email&utm_source=footer
> >.
>
> --
> 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+uns... at googlegroups.com<mailto:cp2k+uns... at googlegroups.com
> >.
> To view this discussion on the web visit 
> https://groups.google.com/d/msgid/cp2k/e7995875-2d28-41ca-8eaa-d69146b28775n%40googlegroups.com
> <
> https://groups.google.com/d/msgid/cp2k/e7995875-2d28-41ca-8eaa-d69146b28775n%40googlegroups.com?utm_medium=email&utm_source=footer
> >.
>
> --
> 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+uns... at googlegroups.com<mailto:cp2k+uns... at googlegroups.com
> >.
> To view this discussion on the web visit 
> https://groups.google.com/d/msgid/cp2k/0b51bf55-96d9-4f6a-aa3a-3f2a77e6f5afn%40googlegroups.com
> <
> https://groups.google.com/d/msgid/cp2k/0b51bf55-96d9-4f6a-aa3a-3f2a77e6f5afn%40googlegroups.com?utm_medium=email&utm_source=footer
> >.
>

-- 
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/a592b851-b53c-4b90-9578-65ecf7a0ae95n%40googlegroups.com.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20230705/acb39e83/attachment-0001.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: pre-compiled binary.png
Type: image/png
Size: 26614 bytes
Desc: not available
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20230705/acb39e83/attachment-0001.png>