Menu
▾
▴
Re: [cclib-users] gbasis with Orca
|
From: Karol L. <kar...@gm...> - 2016-03-04 18:50:09
|
Great! If the tests pass, then we can be quite certain it works correctly. On Fri, Mar 4, 2016 at 9:25 AM, Felix Plasser <fel...@un...> wrote: > Hi, I double checked this and it looks like gbasis is parsed correctly. If > I do the same calculation with Orca and Gaussian, then I get the same > results with > ccget orca.out gbasis > and > ccget gaussian.log gbasis > So everything looks fine and I get the same canonical result. > > The thing that confused me initially is that Orca uses a different order > for the p-functions - and because of that things looked differently. But > this is already parsed correctly by cclib: > ccget orca.out aonames > [ ..., u'C1_1PZ', u'C1_1PX', u'C1_1PY', ...] > And I will just have to reorder the MO-coefficients accordingly for my > purposes. > > As I see it, everything looks fine, and I'll send the pull request. > > By the way, getting the correct ordering of the basis functions would also > be a challenge for the Molden writer. My own little tool cc2molden.py > https://sourceforge.net/p/theodore-qc/code/ci/master/tree/bin/cc2molden.py > failed because of this ordering. And that is what confused me ... > > -Felix > > > On 03/04/2016 03:06 PM, Karol Langner wrote: > > I think we want to parse the canonical gbasis before any normalization, > because that's transferable. In any case, we want it to be the same for all > programs. Is there an easy way to unnormalize? > On Mar 4, 2016 1:49 AM, "Felix Plasser" <fel...@un...> > wrote: > >> Hi Karol and Eric, >> >> I do have the parser now >> >> https://github.com/felixplasser/cclib/commit/0f8e6db49910aedb92a1cad5e0ba384a3fc0efaa?diff=unified >> >> This parses gbasis as it is printed out in the standard output. The >> problem is that internally ORCA changes the normalization factors, and then >> this gbasis is no longer consistent with the MO-coefficients. For example >> when I run orca_2mkl, it tells me >> "The basis set has NOT been un-normalized for printing" >> And then I get the un-normalized coefficients that are used internally >> and that are consistent with mocoeffs. But they are not the same >> coefficients printed in the output file. >> >> For my specific application gbasis works they way it is parsed now. But I >> do not really know what gbasis is supposed to be: the input coefficients, >> or the coefficients used internally. >> >> About Molden format: it would just parse a few attributes, mainly atmnos, >> atomcoords, gbasis, and mocoeffs >> But the important point here would be that gbasis and mocoeffs are >> consistent and well-defined (assuming that the molden file was created >> properly) >> >> Felix >> >> On 03/03/2016 09:35 PM, Eric Berquist wrote: >> >> Hi Felix and Karol, >> >> If a pull request is made, I guarantee I'll take a look at it, since I'm >> trying to be more active in the project. >> >> We don't have a Molden parser yet, though I started a Molden writer for >> cclib, and have something standalone that attempts to clean up the Molden >> output produced by a few different programs, since Molden itself seemed >> pretty picky about the section ordering: >> https://github.com/berquist/scripts/blob/master/fix_molden.py >> >> Eric >> >> On Thu, Mar 3, 2016 at 3:25 PM, Karol Langner <kar...@gm...> >> wrote: >> >>> Yes, pull requests are definitely the way to go. Turnaround time varies >>> with how busy people are. >>> >>> Parsing molden files is possible, just requires some work. However, I'm >>> not sure what kind of coverage of cclib attribute one could get from those >>> files. Are they text format or binary? >>> >>> On Thu, Mar 3, 2016 at 3:01 AM, Felix Plasser < >>> fel...@un...> wrote: >>> >>>> Ok, sounds good. I will try to set it up. And then I'll send a >>>> pull-request via github, if that's the way to do it ... >>>> >>>> By the way, another thing I am thinking about is that cclib would be >>>> able to parse files in Molden format. That would be a nice way to get: >>>> atmonos, atomcoords, gbasis, and mocoeffs in a standardized way. >>>> >>>> Cheers, >>>> Felix >>>> >>>> On 03/03/2016 06:54 AM, Karol Langner wrote: >>>> >>>> Hi Felix, >>>> >>>> If you look at the table showing coverage of attributes ( >>>> http://cclib.github.io/data.html#details-of-current-implementation) >>>> you'll see there's a T/D for gbasis/ORCA. That's automatically generated >>>> from the code, and as you've observed there is no parsing code for that yet >>>> :) apparently I put that note there to not forget. >>>> >>>> If you implement this, please do contribute it to cclib. I probably >>>> won't get around to it anytime soon. >>>> >>>> Cheers, >>>> Karol >>>> >>>> >>>> On Wed, Mar 2, 2016 at 7:05 AM, Karol Langner <kar...@gm...> >>>> wrote: >>>> >>>>> Hi Felix, >>>>> >>>>> It might be lingering in a branch or pull request. Let me check and >>>>> get back to you. >>>>> >>>>> Karol >>>>> On Mar 2, 2016 6:57 AM, "Felix Plasser" <fel...@un...> >>>>> wrote: >>>>> >>>>>> Hi, I am just testing cclib for a new project of ours. >>>>>> >>>>>> What I was wondering is: Is it possible to get the gbasis attribute >>>>>> from Orca? In the "parsed data notes" it says the it should work if I >>>>>> include the line >>>>>> Print[ P_Basis ] 2 in the output block >>>>>> >>>>>> Indeed, that will print out the basis coefficients and exponents (see >>>>>> the attached orca.out file), but the gbasis attribute is not parsed. I >>>>>> checked the newest source from github and it looks like the file >>>>>> orcaparser.py does not contain the word gbasis anywhere. So I assume this >>>>>> is not parsed? >>>>>> >>>>>> I guess it will not be too much trouble for me to add the >>>>>> functionality. I just wanted to check if this is already there before I >>>>>> would try to implement it myself. >>>>>> >>>>>> Thank you, >>>>>> Felix >>>>>> >>>>>> -- >>>>>> Felix Plasser >>>>>> Institute for Theoretical Chemistry >>>>>> University of Vienna >>>>>> Währingerstr. 17/404 >>>>>> 1090 Wien >>>>>> fel...@un... >>>>>> http://homepage.univie.ac.at/felix.plasser/ >>>>>> Tel.: +43-1-4277-52757 >>>>>> >>>>>> >>>>>> >>>>>> ------------------------------------------------------------------------------ >>>>>> Site24x7 APM Insight: Get Deep Visibility into Application Performance >>>>>> APM + Mobile APM + RUM: Monitor 3 App instances at just $35/Month >>>>>> Monitor end-to-end web transactions and take corrective actions now >>>>>> Troubleshoot faster and improve end-user experience. Signup Now! >>>>>> http://pubads.g.doubleclick.net/gampad/clk?id=272487151&iu=/4140 >>>>>> _______________________________________________ >>>>>> cclib-users mailing list >>>>>> ccl...@li... >>>>>> https://lists.sourceforge.net/lists/listinfo/cclib-users >>>>>> >>>>>> >>>> >>>> -- >>>> Felix Plasser >>>> Institute for Theoretical Chemistry >>>> University of Vienna >>>> Währingerstr. 17/404 >>>> 1090 Wie...@un...http://homepage.univie.ac.at/felix.plasser/ >>>> Tel.: +43-1-4277-52757 >>>> >>>> >>>> >>>> ------------------------------------------------------------------------------ >>>> Site24x7 APM Insight: Get Deep Visibility into Application Performance >>>> APM + Mobile APM + RUM: Monitor 3 App instances at just $35/Month >>>> Monitor end-to-end web transactions and take corrective actions now >>>> Troubleshoot faster and improve end-user experience. Signup Now! >>>> http://pubads.g.doubleclick.net/gampad/clk?id=272487151&iu=/4140 >>>> _______________________________________________ >>>> cclib-users mailing list >>>> ccl...@li... >>>> https://lists.sourceforge.net/lists/listinfo/cclib-users >>>> >>>> >>> >>> >>> ------------------------------------------------------------------------------ >>> Site24x7 APM Insight: Get Deep Visibility into Application Performance >>> APM + Mobile APM + RUM: Monitor 3 App instances at just $35/Month >>> Monitor end-to-end web transactions and take corrective actions now >>> Troubleshoot faster and improve end-user experience. Signup Now! >>> http://pubads.g.doubleclick.net/gampad/clk?id=272487151&iu=/4140 >>> _______________________________________________ >>> cclib-users mailing list >>> ccl...@li... >>> https://lists.sourceforge.net/lists/listinfo/cclib-users >>> >>> >> >> -- >> Felix Plasser >> Institute for Theoretical Chemistry >> University of Vienna >> Währingerstr. 17/404 >> 1090 Wie...@un...http://homepage.univie.ac.at/felix.plasser/ >> Tel.: +43-1-4277-52757 >> >> >> >> ------------------------------------------------------------------------------ >> Site24x7 APM Insight: Get Deep Visibility into Application Performance >> APM + Mobile APM + RUM: Monitor 3 App instances at just $35/Month >> Monitor end-to-end web transactions and take corrective actions now >> Troubleshoot faster and improve end-user experience. Signup Now! >> http://pubads.g.doubleclick.net/gampad/clk?id=272487151&iu=/4140 >> _______________________________________________ >> cclib-users mailing list >> ccl...@li... >> https://lists.sourceforge.net/lists/listinfo/cclib-users >> >> > -- > Felix Plasser > Institute for Theoretical Chemistry > University of Vienna > Währingerstr. 17/404 > 1090 Wie...@un...http://homepage.univie.ac.at/felix.plasser/ > Tel.: +43-1-4277-52757 > > > > ------------------------------------------------------------------------------ > > _______________________________________________ > cclib-users mailing list > ccl...@li... > https://lists.sourceforge.net/lists/listinfo/cclib-users > > |
