CamCASP/Bugs/3

• Differences in results with and without rotation of integrals

I have noticed that OVERLAP integrals (water dimer vtz/vtz) calculated by rotation of integrals (REDO-DF-ON-ROTATION false) differ by about 1% or less from integrals calculated by repeating the density-fitting.

The differences in the total overlap seem to vanish if the molecules are not translated. So does that mean that the integrals calculated in CamCASP are not translationally invariant?

The distributed overlap integrals differ in all cases by 1% or less. --alston 19:17, 30 October 2008 (GMT)

FIXED.

Integrals with the auxiliary basis functions were not being normalized correctly when spherical GTOs were used. There was a spurious factor of

<math> \frac{1}{\sqrt{(2i-1)!!(2j-1)!!(2k-1)!!}} </math>

that was being retained when it should not have been there. GAMINT includes such a factor which was correctly removed from integrals involving the main basis functions but not the auxiliary basis functions. Well, this is now taken care of.

This error didn't effect any earlier calculations as all auxiliary basis functions had the wrong normalization. It was only when I attempted to operate with the Wigner matrices that the problem showed up. These matrices are constructed to rotate normalized spherical functions. So the rotation went wrong. The errors were small because they appeared only in the d and higher symmetries which were not so important for the water dimer example I used. Particularly in the vdz basis! --alston 14:32, 4 November 2008 (GMT)