The method used by this program is quite similar to MOKIT. However, we try to do everything with PySCF and without Gaussian.
This program aims to do:
- HF guess strategy
- automatic guess for CASSCF/GVB/SUHF
- interface for post-MR
Pre-requisites
- MOKIT (no need to fully compile, only
autopair
is needed) - PySCF
- mrh (optional, for MC-PDFT)
- ExSCF (optional, for SUHF)
- pyNOF (optional, for GVB)
Install
- git clone and add
/path/to/pyAutoMR
to yourPYTHONPATH
- UHF -> UNO (-> PM LMO -> assoc rot) (-> GVB) -> CASSCF
- UHF -> SUHF -> CASSCF
- RHF (-> vir MO projection -> PM LMO -> pairing) (-> GVB ) -> CASSCF
- CASSCF -> MC-PDFT
- CASSCF(dry run) -> SA-CASSCF
UHF, RHF can be auto-detected.
- guess for UHF/UKS
- mix
- fragment
- from_fch
- flipspin (by lmo or by site)
- internal stability for RHF/RKS, UHF/UKS, ROHF/ROKS
- optimize wavefunction until stable
- dump CASCI coefficients
- dump (active) orbital compositions
from automr import guess, autocas
xyz = 'N 0.0 0.0 0.0; N 0.0 0.0 1.9'
bas = 'cc-pvdz'
mf = guess.from_frag(xyz, bas, [[0],[1]], [0,0], [3,-3], cycle=50)
mf = guess.check_stab(mf)
mf2 = autocas.cas(mf)
- TDDFT NO -> CASSCF
Please cite pyAutoMR as
Shirong Wang, pyAutoMR, https://github.com/jeanwsr/pyAutoMR (accessed month day, year)
and cite every program called by pyAutoMR, such as PySCF, MOKIT, mrh, etc.