Unchaining a Raman calculaiton

Phonons, DFPT, electron-phonon, electric-field response, mechanical response…

Moderators: mverstra, joaocarloscabreu

Locked
User avatar
jzwanzig
Posts: 504
Joined: Mon Aug 17, 2009 9:25 am

Unchaining a Raman calculaiton

Post by jzwanzig » Mon Apr 18, 2011 10:52 am

Hi,
the NLO tutorial explains how to do a Raman calculation using chained datasets, but it would be convenient also to do it with unchained sets. The problem seems to be the lack of the input variable "ird1den". Note that in the chained case, you need getwfk, get1wfk, getden, and get1den, and there are "ird" counterparts of all of these except "ird1den". Is there a deep reason for this lack or is it something that I could add?

thanks,
Joe
Josef W. Zwanziger
Professor, Department of Chemistry
Canada Research Chair in NMR Studies of Materials
Dalhousie University
Halifax, NS B3H 4J3 Canada
jzwanzig@gmail.com

User avatar
jzwanzig
Posts: 504
Joined: Mon Aug 17, 2009 9:25 am

Re: Unchaining a Raman calculaiton

Post by jzwanzig » Wed May 04, 2011 7:13 pm

Well, as no one answered my original query, I went ahead and added the input variable ird1den. I probably don't understand MV's earlier post that "ird1den is implicit"; in the test file tnlo_2.in, if you comment out get1den in the last dataset the calculation doesn't work, so it does not seem to be looking for the 1den datasets it needs, at least in my hands. Be that as it may, I've added the variable ird1den, documented it, and added into the loop3dte file and associated references so now raman calculations can be done piecewise, in addition to chaining them together. I have not figured out how to test ird1den in the test suite yet (I notice we are not testing ird1wf either, once I figure this out I will test both in a single pass).
Josef W. Zwanziger
Professor, Department of Chemistry
Canada Research Chair in NMR Studies of Materials
Dalhousie University
Halifax, NS B3H 4J3 Canada
jzwanzig@gmail.com

User avatar
jzwanzig
Posts: 504
Joined: Mon Aug 17, 2009 9:25 am

Re: Unchaining a Raman calculaiton

Post by jzwanzig » Sun May 08, 2011 12:27 pm

Now in my current branch (6.6.3-public) I have added tests for both ird1wf and ird1den, when 6.6.3 is released this will be available to all.
Josef W. Zwanziger
Professor, Department of Chemistry
Canada Research Chair in NMR Studies of Materials
Dalhousie University
Halifax, NS B3H 4J3 Canada
jzwanzig@gmail.com

mverstra
Posts: 655
Joined: Wed Aug 19, 2009 12:01 pm

Re: Unchaining a Raman calculaiton

Post by mverstra » Thu May 12, 2011 8:50 pm

Hi Joe,

excellent news - we leave you alone for a month and you answer your own questions, code what's missing - what more?

I have also worried about restarting RF calculations (eg with a different k grid, or a density gotten elsewhere) but always got around this by making a slew of datasets, and using get1den if needed, selecting just the ones I wanted to do with jdtset.

By implicit in the old post I just meant that the 1DEN is read in anyway (with get1den) and necessarily, in order to do the 3rd order stuff.

The trouble you may be having (stop me if I insult) is that the ird1den is looking for a slightly different file name (example for running DS4)

inp_DS4_1DEN

instead of

out_DS3_1DEN

The former is obviously not created yet at the beginning of DS4 when it needs to be read in. You'd have to copy the file by hand - abinit will never be able to do this. Did you post the input file you were using?

ciao

Matthieu
Matthieu Verstraete
University of Liege, Belgium

Locked