MolSSI / QCEngine

Quantum chemistry program executor and IO standardizer (QCSchema).

https://molssi.github.io/QCEngine/

BSD 3-Clause "New" or "Revised" License

163 stars 79 forks source link

slightly loosen xtb testing #308

Closed loriab closed 3 years ago

loriab commented 3 years ago

Description

will this be ok, @awvwgk? it fixes CI.

Status

[x] Code base linted
[x] Ready to go

codecov[bot] commented 3 years ago

Codecov Report

Merging #308 (ca8f8c6) into master (cf5fdc5) will not change coverage. The diff coverage is 100.00%.

loriab commented 3 years ago

will this be ok, @awvwgk? it fixes CI. (sorry if double-ping -- I don't think GH pings on edits.)

awvwgk commented 3 years ago

I think the current threshold is relative rather than absolute (https://docs.pytest.org/en/6.2.x/reference.html#pytest.approx). I switched to pytest.approx(..., abs=thr) for testing in my projects since the relative comparison can fail for very small (numerical zero) values otherwise.

awvwgk commented 3 years ago

sorry if double-ping -- I don't think GH pings on edits.

No problem. Looks like GH pings correctly for mentions in edits.

loriab commented 3 years ago

I think the current threshold is relative rather than absolute (https://docs.pytest.org/en/6.2.x/reference.html#pytest.approx). I switched to pytest.approx(..., abs=thr) for testing in my projects since the relative comparison can fail for very small (numerical zero) values otherwise.

Sounds good -- it was checking some elements to 1e-11 before, so this is a sensible relaxation.