Add support for ORCA

[DetlevCM]

Assuming work is done on other codes, adding support for ORCA could be interesting.

The code is both fast, offers many good new methods that will be interesting (e.g. look at Paulechka and Kazakov 2017 who used DLPNO) and has excellently structured output.

[mattasplund]

It would be. Conversion of the internal ECCE data structures to input files and from output files back to ECCE is all in easily editable scripts, but the documentations is not wonderful. Could be a great student project, though.

On Oct 2, 2018, at 11:14 AM, Detlev C M ***@***.***> wrote: Assuming work is done on other codes, adding support for ORCA could be interesting. The code is both fast, offers many good new methods that will be interesting (e.g. look at Paulechka and Kazakov 2017 who used DLPNO) and has excellently structured output. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#38>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AcDHgKLwO5cgSQmS5j9Bnbr1A3U0ecLlks5ug57dgaJpZM4XEfBa>.

Matthew Asplund matt@asplundfamily.org

[ohlincha]

I think the key here is that someone who

1. is a proficient user of ORCA
2. has a strong professional interest in using ECCE with ORCA
   needs to undertake the implementation.

If anyone does this and submits the code to us we'd be very happy to include it.

[DetlevCM]

Using ORCA is very easy/trivial - at least for "standard computational chemistry". If you want to do say a CAS calculation, of course it is more involved. The "basics" are possibly amongst the most accessible of any computational chemistry program.
One huge advantage of ORCA is, that the output is very user-friendly.
I would call it the best I have seen.

Interest can be generated if it is supported at least to a certain minimum. - Obviously if there is no support, there will be no interest.

What kind of scripts does ECCE use? I fiddled with ECCE back when I was in the UK, testing NWChem a bit - but I didn't delve anywhere into the code/scripts.
Maybe I can help a bit?

I think as a "minimum support" it would be nice to:

1. create input files for optimisations and frequencies
2. read the resulting geometries and energies and frequencies from the output

These two would be fairly trivial and would allow for determining if there is interest in additional features.

[mattasplund]

There is a set of python scripts that describe how to create an input file. I played around a bit, and it didn’t take much to take a copy of one of the other python scripts, copy them, and then start editing. I don’t know much about ORCA, but it should be pretty easy to get started. In the apps/data/client/cap directory there are a series of .edml files that describe the capabilities of each compute code, so you would just have to make a copy from gaussian or nwchem and start editing. Then under apps/scripts/codereg there are files that tell what information should show up in the runtype and theory dialog boxes. Again, make a copy of an existing code and start editing. Finally, there are a set of perl scripts and templates in apps/scripts/parsers that describe how to make an input file, and how to interpret the output. I am definitely not a perl hacker, so that is a bit out of my experience. The really good news is that there is no hacking of the binary necessary to implement it. Their design moved all of the interaction with outside code into the python and perl scripts specifically to make this sort of thing easy (or at least easier). Matthew Asplund

On Oct 4, 2018, at 5:47 AM, Detlev C M ***@***.***> wrote: Using ORCA is very easy/trivial - at least for "standard computational chemistry". If you want to do say a CAS calculation, of course it is more involved. The "basics" are possibly amongst the most accessible of any computational chemistry program. One huge advantage of ORCA is, that the output is very user-friendly. I would call it the best I have seen. Interest can be generated if it is supported at least to a certain minimum. - Obviously if there is no support, there will be no interest. What kind of scripts does ECCE use? I fiddled with ECCE back when I was in the UK, testing NWChem a bit - but I didn't delve anywhere into the code/scripts. Maybe I can help a bit? I think as a "minimum support" it would be nice to: create input files for optimisations and frequencies read the resulting geometries and energies and frequencies from the output These two would be fairly trivial and would allow for determining if there is interest in additional features. — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#38 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AcDHgIAX5m8M-CwD8AIgYxjWpkPFw36iks5uhfU7gaJpZM4XEfBa>.

Matthew Asplund matt@asplundfamily.org

[mattasplund]

OH, and there is a set of slides from Gary Black that you can see at http://ecce.emsl.pnl.gov/docs/code_reg_slides.pdf <http://ecce.emsl.pnl.gov/docs/code_reg_slides.pdf>. Sadly Gary has been forced to move to new projects, but he does try to help where he can.

On Oct 4, 2018, at 5:47 AM, Detlev C M ***@***.***> wrote: Using ORCA is very easy/trivial - at least for "standard computational chemistry". If you want to do say a CAS calculation, of course it is more involved. The "basics" are possibly amongst the most accessible of any computational chemistry program. One huge advantage of ORCA is, that the output is very user-friendly. I would call it the best I have seen. Interest can be generated if it is supported at least to a certain minimum. - Obviously if there is no support, there will be no interest. What kind of scripts does ECCE use? I fiddled with ECCE back when I was in the UK, testing NWChem a bit - but I didn't delve anywhere into the code/scripts. Maybe I can help a bit? I think as a "minimum support" it would be nice to: create input files for optimisations and frequencies read the resulting geometries and energies and frequencies from the output These two would be fairly trivial and would allow for determining if there is interest in additional features. — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#38 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AcDHgIAX5m8M-CwD8AIgYxjWpkPFw36iks5uhfU7gaJpZM4XEfBa>.

Matthew Asplund matt@asplundfamily.org

[DetlevCM]

Thank you for the information.

I really do not like Python and I don't know Perl, but I will try to get some basic functionality to work to support ORCA (input files first I guess as they are the easiest).