exalearn / ExaMol / 6481915426

98%

push

github

Tweak the performance of redox, gather more data (#115)

* Improve the parallelism settings

* Expand cell more aggressively

* Switch from QN to BFGS

* Account for cell changes

Our script now re-adjusts cell size, which leads to multiple records for the same geometry. This is OK because I would like to store multiple measurements for the same geometry so that  we can establish error bounds on the forces/energies wrt cell size.

* Update runs with new data

* Flake8 fix

7 of 7 new or added lines in 1 file covered. (100.0%)

1786 of 1818 relevant lines covered (98.24%)

0.98 hits per line