Andy Jiang

> How does the solid harmonics code different from the solid harmonics code found in libmints? https://github.com/psi4/psi4/blob/master/psi4/src/psi4/libmints/solidharmonics.cc My code here uses the Helgaker multipole normalization convention, and it introduces the...

> > How does the solid harmonics code different from the solid harmonics code found in libmints? https://github.com/psi4/psi4/blob/master/psi4/src/psi4/libmints/solidharmonics.cc > > My code here uses the Helgaker multipole normalization convention, and...

> Fair point if the solidharmonics.cc code is outdated. I was curious if you were re-inventing something we already have. > > @JonathonMisiewicz I'll hold off commenting on whether a...

> > Iirc, @andysim said that the old libmints solidharmonics.cc code is outdated anyways, and the Helgaker recursions are better. Feel free to comment Andy. > > Both codes use...

@jturney @JonathonMisiewicz @andysim It appears that the current solidharmonics.cc is not used ANYWHERE in the Psi4 code. I am actually in favor of moving all my fmm stuff to libmints....

> > It appears that the current solidharmonics.cc is not used ANYWHERE in the Psi4 code. > > Somehow solidharmonics.cc is getting used, because changing it changes the integral ordering...

> My vote is to either end or temporarily pause this PR and open a new PR just for swapping out the solid harmonics technology. Of course, @jturney's judgment overrides...

@davpoolechem Since #3098 has been opened, should we close this PR?

Just implemented some of Lori and Tibor's suggestions. I have also implemented SC-LMP2 for "weak pairs" to reduce the cost of the LCCSD computation, per the Sparse Maps II paper.

Requesting additional reviews for this PR. @jturney, @susilehtola?