CI seems to be failing due to bad data collection in the test_gene_expression test in test_integration.py.
Failures seem to happen at random across multiple OSes and Python versions. First noticed in #320, but then attempted to run on a branch identical to main and got similar results in #321. I would guess it's some sort of connection problem since some of #320's tests managed to pass the second time round, but not the first. Python 3.9 and 3.10 consistently fail at different points however.
I cannot replicate this bug by running the tests locally through tox on my machine, nor by performing the test steps manually in a Python terminal. So either it's the GitHub runners or there's a configuration on my work PC.
As an aside, this repo's test workflow does not have a workflow_dispatch trigger which makes debugging of this kind very hard (requires one to open a pull request and push to it to trigger any sort of testing). This feature is added in #320, once we get to the bottom of this.
CI seems to be failing due to bad data collection in the
test_gene_expression
test intest_integration.py
.Failures seem to happen at random across multiple OSes and Python versions. First noticed in #320, but then attempted to run on a branch identical to main and got similar results in #321. I would guess it's some sort of connection problem since some of #320's tests managed to pass the second time round, but not the first. Python 3.9 and 3.10 consistently fail at different points however.
I cannot replicate this bug by running the tests locally through tox on my machine, nor by performing the test steps manually in a Python terminal. So either it's the GitHub runners or there's a configuration on my work PC.
As an aside, this repo's test workflow does not have a
workflow_dispatch
trigger which makes debugging of this kind very hard (requires one to open a pull request and push to it to trigger any sort of testing). This feature is added in #320, once we get to the bottom of this.