Stuck on an issue?

Lightrun Answers was designed to reduce the constant googling that comes with debugging 3rd party libraries. It collects links to all the places you might be looking at while hunting down a tough bug.

And, if you’re still stuck at the end, we’re happy to hop on a call to see how we can help out.

tests: Fix SciPy builds to run nightly tests against SciPy HEAD

See original GitHub issue

Description

At the moment testing against the HEAD of SciPy’s GitHub repo is causing timeout failures and isn’t helping to catch anything

https://github.com/scikit-hep/pyhf/blob/025ffe9be212f53a8eb6c46f7a53cc3e50017d81/.github/workflows/dependencies-head.yml#L24-L31

This is being reverted to just using --pre to test against release candidates, but it would be good to figure out how to get a working build.

Trying to follow what SciPy is doing in GHA for build tests doesn’t seem to actually install it.

All the testing on this is being done on the fix/get-scipy-building dev branch.

Issue Analytics

State:
Created 2 years ago
Comments:6 (4 by maintainers)

Top GitHub Comments

1reaction

matthewfeickertcommented, Jul 27, 2021

The build log is hidden, that’s why it’s hard to see what is going on.

This probably isn’t a huge help at the moment, but here’s a .zip file with the logs for a recent build of SciPy on the test branch that I canceled after an hour:

scipy-ubuntu-latest-python3.8.zip

I can make the logs more verbose in the future.

1reaction

rgommerscommented, Jul 27, 2021

A few questions/suggestions:

Do you need to build SciPy from source? There are nightly wheels at https://anaconda.org/scipy-wheels-nightly/
Looking at dependencies_head.yml, you are now installing both ATLAS and plain LAPACK. Why the latter? Maybe it happens to pick up Netlib LAPACK for some reason rather than failing, and that’s what makes it slow? Is your test suite heavy on linear algebra?