2tRDM between different HCI states

[ghb24]

Slightly non-standard usage, but I was wondering whether the following workflow would be something possible within this code.

I am hoping to compute two HCI states of two different hamiltonians (e.g. with and without an explicit electric field), and then compute the overlap and 2-transition-RDMs between them. I know that HCI has the capability to compute 2-transition-RDMs, but these are by default between the excited states within a single run. Is it possible to have the code e.g. read back in an HCI state, and then compute the 2tRDMs between these states (where they have been optimized with different hamiltonians, though in the same basis)? Similarly, it would also be good to compute the overlap between these states, which again I can't see a functionality to do out of the box. I am obviously fine with just dealing with the variational component of these states in this usage.

Many thanks for any help with this request.

[sanshar]

The basis set and geometry is exactly the same? In this case one will have to hack the c++ code a bit and get it to read an optimized wavefunction from a previous calculation and calculate the transition RDM with the state calculated in the current run. Currently I don't think there are options to do it without some elbow grease. If you are optimizing the orbitals using two independent CASSCF runs then it is more challenging because the basis in which one is calculating the HCI state changes. Sandeep.

…

On Tue, May 14, 2024 at 10:35 AM ghb24 ***@***.***> wrote: Slightly non-standard usage, but I was wondering whether the following workflow would be something possible within this code. I am hoping to compute two HCI states of two different hamiltonians (e.g. with and without an explicit electric field), and then compute the overlap and 2-transition-RDMs between them. I know that HCI has the capability to compute 2-transition-RDMs, but these are by default between the excited states within a single run. Is it possible to have the code e.g. read back in an HCI state, and then compute the 2tRDMs between these states (where they have been optimized with different hamiltonians, though in the same basis)? Similarly, it would also be good to compute the overlap between these states, which again I can't see a functionality to do out of the box. I am obviously fine with just dealing with the variational component of these states in this usage. Many thanks for any help with this request. — Reply to this email directly, view it on GitHub <#10>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AABVW4DATXLCWIVJYLYWMFDZCI4MVAVCNFSM6AAAAABHWQ562OVHI2DSMVQWIX3LMV43ASLTON2WKOZSGI4TKOJRGYZTQNA> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

[ghb24]

Thanks for getting back to me so quickly Sandeep.

Yes, while the orbitals will be the same, and just the Hamiltonian different, I did expect it would require a little hacking around in the code to get it to work. I might follow it up with you offline.

George

[sanshar]

Okay, I can hack something together for you pretty quickly. Just send me an example hamiltonian. Sandeep.

…

On Tue, May 14, 2024 at 2:19 PM ghb24 ***@***.***> wrote: Thanks for getting back to me so quickly Sandeep. Yes, while the orbitals will be the same, and just the Hamiltonian different, I did expect it would require a little hacking around in the code to get it to work. I might follow it up with you offline. George — Reply to this email directly, view it on GitHub <#10 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AABVW4GP3XRTL64WI4VKEQTZCJWWTAVCNFSM6AAAAABHWQ562OVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDCMJRGA3TEMJYGQ> . You are receiving this because you commented.Message ID: ***@***.***>