Implement imaginary time-propagation for all solvers

[Schoyen]

In an attempt to find the (actual) OACCD ground state, we can try to implement an imaginary time-propagation scheme for the already existing time-dependent solvers. This would benefit from the change of convention of the time-dependent solvers as proposed by @halvarsu.

[tbpedersen]

Before embarking on this, we should talk to Rolf Myhre at NTNU. He started working on this (in collaboration with Simen and myself), but I am not updated on the current status. At any rate, there’s no reason to start from scratch.

Thomas Bondo Pedersen, Professor Hylleraas Centre for Quantum Molecular Sciences Department of Chemistry, University of Oslo Email: t.b.pedersen@kjemi.uio.nomailto:t.b.pedersen@kjemi.uio.no Phone: +47 22857182 Mobile: +47 94978990

On 4 Feb 2020, at 17:33, Øyvind Sigmundson Schøyen notifications@github.com<mailto:notifications@github.com> wrote:

In an attempt to find the (actual) OACCD ground state, we can try to implement an imaginary time-propagation scheme for the already existing time-dependent solvers. This would benefit from the change of convention of the time-dependent solvers as proposed by @halvarsuhttps://github.com/halvarsu.

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHubhttps://github.com/Schoyen/coupled-cluster/issues/26?email_source=notifications&email_token=ABB5SNRTRQPOMLHOE5YRS5TRBGKFJA5CNFSM4KP2A2SKYY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4IK6NXFQ, or unsubscribehttps://github.com/notifications/unsubscribe-auth/ABB5SNQAQD26DDK3JMDAMN3RBGKFJANCNFSM4KP2A2SA.

[rolfheil]

Hi, I did some stuff with this and it's probable somewhere on the repository.

It worked for CCSD, but not particularly well.

On 04.02.2020 21:08, Thomas Bondo Pedersen wrote:

Before embarking on this, we should talk to Rolf Myhre at NTNU. He started working on this (in collaboration with Simen and myself), but I am not updated on the current status. At any rate, there’s no reason to start from scratch.

Thomas Bondo Pedersen, Professor Hylleraas Centre for Quantum Molecular Sciences Department of Chemistry, University of Oslo Email: t.b.pedersen@kjemi.uio.nomailto:t.b.pedersen@kjemi.uio.no Phone: +47 22857182 Mobile: +47 94978990

On 4 Feb 2020, at 17:33, Øyvind Sigmundson Schøyen notifications@github.com<mailto:notifications@github.com> wrote:

In an attempt to find the (actual) OACCD ground state, we can try to implement an imaginary time-propagation scheme for the already existing time-dependent solvers. This would benefit from the change of convention of the time-dependent solvers as proposed by @halvarsuhttps://github.com/halvarsu.

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHubhttps://github.com/Schoyen/coupled-cluster/issues/26?email_source=notifications&email_token=ABB5SNRTRQPOMLHOE5YRS5TRBGKFJA5CNFSM4KP2A2SKYY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4IK6NXFQ, or unsubscribehttps://github.com/notifications/unsubscribe-auth/ABB5SNQAQD26DDK3JMDAMN3RBGKFJANCNFSM4KP2A2SA.

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/Schoyen/coupled-cluster/issues/26?email_source=notifications&email_token=ABRJTXG4IU4IROY67FW6THLRBHDMXA5CNFSM4KP2A2SKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEKZAKWY#issuecomment-582092123, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABRJTXEDV5OL3COGIOF5QETRBHDMXANCNFSM4KP2A2SA.

[Schoyen]

Good idea! We'll have a look at your implementation @rolfheil!

[haakoek]

There is a discussion of imaginary time propagation for time-dependent vibrational coupled-cluster (same underlying amplitude equations) in section H of this article: https://aip.scitation.org/doi/10.1063/1.5117207