Restructure determinant treatment before gtdmctl loop in RASSI
Change the structure of core RASSI
routine gtdmctl
:
- CI expansion in Slater determinant basis is not written anymore to the disk for one of the JOBs, both wave function decompositions are now saved to memory for each pair of JOBs
- Print out and save to HDF5 file (new
CIH5
keyword) of determinant expansion and determinant occupation vectors are organized, which is needed for external SCAMPI program and generally useful to have - There is commented code added to execute the
gtdmctl
loop (over pairs of states) in parallel. It is working only for simplest case, when oscillator strengths are needed. I struggled to make it work for general case, but there are a lot of places inside this loop I am not really aware of. I hope it is ok to leave this commented code, might be useful eventually for parallelization of the loop in the future.
Small changes while testing compatibility with the new code in RASSI
:
readme.md
file in auger-oca
tools. @brncat please check whether everything is displayed correctly.
-
Documentation updated or not needed. -
Test created or not needed. -
Attribution in source files and CONTRIBUTORS.md
updated if necessary.
Edited by Vladislav Kochetov