Molcas Forum

andrewshyichuk

Member

Registered: 2020-02-13

Posts: 86

Slapaf and me cannot agree on things.

Dear Users,

I want to get a gradient and the first geom opt coordinates for a system.

My input goes as: &SEWARD, &SCF, &SLAPAF.

This works. But then, if I do another SCF, or a RASSCF - SLAPAF does not want to do it's thing any more, with an error:

 ****************** ERROR *********************
 Coordinates did not change!
 Maybe SEWARD is not inside the loop?
 **********************************************

Well, yes, SEWARD is not in a loop for the lack of loop.

I assume the Runfile contain information about a previous run of Slapaf. Any way to ignore it?

I can redo the integrals but those take like 6 hours, and also I find it a stupid fix, sorry.

Thank you.

Ignacio

Administrator

From: Uppsala

Registered: 2015-11-03

Posts: 1,085

Re: Slapaf and me cannot agree on things.

Well, yes. SLAPAF saves information into the runfile for the next iteration. If the next "iteration" is not what it expects (i.e., the new geometry), it will get confused. I woud try saving the runfile before the first SLAPAF run and restoring it after.

If you just want the gradient, run ALASKA instead.

andrewshyichuk

Member

Registered: 2020-02-13

Posts: 86

Re: Slapaf and me cannot agree on things.

Thank you.
Saving runfile... a fix could not have been more simple than this, and yet I did not think of it. I had the correct reason!
That is a kind of moment I'm genuinely happy and grateful for getting help.

Last edited by andrewshyichuk (2024-12-31 12:32:16)