I am trying to extract the two-electron integrals in AOs. I followed the discussion in the forum: http://iopenshell.usc.edu/forum/topic.php?id=3091

I save the binary files 21.0 and 801.0. The file 21.0 keeps all significant integrals and the binary file 801.0 keeps the indices. So far so well, but I noticed that some indices groups (i,j | k,m) are repeating and the corresponding integral values are identical.

Can you please explain this? Is there an underlying ordering convention for the indices i, j and k,m?

(I assume the AO integral indices (i,j | k,m) corresponds to : i,j for the first electron and k,m for the second electron
e.g. (i,j | k,m) = Integral { dr1 dr2 Φi(1)Φj(1)[1/r] Φk(2)Φm(2) } )

I am attaching my input file and a part of the replicated integrals:

$ rem
EXCHANGE HF
BASIS 3-21G*
SCF_CONVERGENCE 8
SCF_GUESS CORE
MAXSCF 2000
SAVE_AO_INTEGRALS 2
USE_NEW_PATH2 FALSE
SYMMETRY FALSE
PURECART 22222
$ end

2 2 2 2 1.21E+00
2 2 3 3 1.20E+00
2 2 4 4 1.20E+00
2 2 5 5 1.20E+00
3 2 3 2 2.72E-01
3 2 3 2 2.72E-01
4 2 4 2 2.72E-01
4 2 4 2 2.72E-01
5 2 5 2 2.72E-01
5 2 5 2 2.72E-01
3 2 3 2 2.72E-01
3 2 3 2 2.72E-01
3 3 2 2 1.20E+00
« Last edit by panant on Wed Apr 06, 2016 10:02 am. »