[molpro-user] CPMCSCF - no. iterations and sourcecode changes

Sve N goretoffel at hotmail.de
Thu Feb 7 16:06:49 GMT 2013


Dear molpro users and developers,
when calculating gradients or, more often non-adiabatic couplings with multi, the routine sometimes does not converge at first, but when the whole calculation is restarted. I assume, there is some information about the previous run saved on file 2, which is used at the beginning of the next try. If I'm not mistaken, one could change the maximum number of internal restarts in line 349 in mucpnr.F, as well as the iterationnumber per restart in form of "maxdim" in line 17.

My main question is, is this a big enough change that I need special permission to play with both values and whom must I contact?

The second, more out of curiosity, what's the reason for both values being hardcoded and why are there two nested loops - is there a different effect to be expected from changing either one, or the other? (maxdim will cost more resources, as I see it, but it won't restart the minimization, so maybe it is the more effective, but overall slowing down switch...)

Best regards, Sven
 		 	   		  
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.molpro.net/pipermail/molpro-user/attachments/20130207/912b2eda/attachment.html>


More information about the Molpro-user mailing list