Changeset 333e84 for pcp/src

Timestamp:

Apr 21, 2008, 2:51:03 PM (17 years ago)

Author:

Frederik Heber <heber@…>

Children:

1da479

Parents:

515271

Message:

CalculateForce(): FIX - P0-perturbation took longer and results were not the same compared to initial commit

There were actually two things that came out to have the same cause:
-P0 minimisation took longer than P1,P2,RxP0,...,RxP2 (especially even against the latter)
-minimisation results, i.e. shielding was not the same anymore and in my c2h4 test setting, the two 13C-shielding had a greater mismatch (2.1539 <-> 2.1507 and later 2.12 <-> 2.153)
The cause was the commented-out line PsiStep=MaxPsiStep, which was clearly nonsense and had me puzzled even on checkin

File:

• pcp/src/run.c (modified) (1 diff)

pcp/src/run.c

@@ -1080 +1080 @@
     while ((R->LevSNo > R->InitLevSNo) || (!Stop && R->LevSNo == R->InitLevSNo)) {
       // occupied
-      //R->PsiStep = R->MaxPsiStep; // reset in-Psi-minimisation-counter, so that we really advance to the next wave function
+      R->PsiStep = R->MaxPsiStep; // reset in-Psi-minimisation-counter, so that we really advance to the next wave function
       R->OldActualLocalPsiNo = R->ActualLocalPsiNo; // reset OldActualLocalPsiNo, as it might still point to a perturbed wave function from last level
       UpdateGramSchOldActualPsiNo(P,Psi);