EDA Functionality for LASSI-PDFT

[JangidBhavnesh]

I have added / pointed the analyze and EDA functionality.

Related to #120

[MatthewRHermes]

If pyscf-forge #64 is accepted, will it be possible to simplify this? And can you add a unit test for it?

[JangidBhavnesh]

If pyscf-forge #64 is accepted, will it be possible to simplify this? And can you add a unit test for it?

I have addressed these. Currently, I have to keep the duplicate of the function from the mcpdft to adopt the nroots as a list.

[MatthewRHermes]

If pyscf-forge #64 is accepted, will it be possible to simplify this? And can you add a unit test for it?

I have addressed these. Currently, I have to keep the duplicate of the function from the mcpdft to adopt the nroots as a list.

OK but ideally, we wouldn't change the datatype of mc.fcisolver.nroots because that may cause many versions of this problem down the road. A cleaner solution imo is to set mc.fcisolver.nroots to the length of the state list, store the state list somewhere else, and modify make_one_casdm* to index into the state list.

ETA: But I won't enforce this before accepting, just address my other comment about identifying sequences.

[JangidBhavnesh]

If pyscf-forge #64 is accepted, will it be possible to simplify this? And can you add a unit test for it?

I have addressed these. Currently, I have to keep the duplicate of the function from the mcpdft to adopt the nroots as a list.

OK but ideally, we wouldn't change the datatype of mc.fcisolver.nroots because that may cause many versions of this problem down the road. A cleaner solution imo is to set mc.fcisolver.nroots to the length of the state list, store the state list somewhere else, and modify make_one_casdm* to index into the state list.

ETA: But I won't enforce this before accepting, just address my other comment about identifying sequences.

Yes. I will make these changes before you accept this PR.

[MatthewRHermes]

Thanks!