[AG-1105] Data-Driven Test for transform_genes_biodomains

[BWMac]

This PR implements the TestTransformGenesBiodomains class in transform_genes_biodomains.py. The new class is made up of some attribute data objects used for the parameterized testing, and two test functions, test_transform_genes_biodomains_should_pass and test_transform_genes_biodomains_should_fail. Each test imports the example input files, attempts to run genes_biodomains.transform_genes_biodomains, and checks the result. In the passing case, the resulting output_df is checked against the expected_output which is provided by the example output files. In the failing case, we expect the function call to result in a KeyError which is verified.

The parameterized testing runs are also annotated with IDs that describe which condition is being tested. When you run

pytest tests/transform/test_genes_biodomains.py -v

the output from the new tests looks like

tests/transform/test_genes_biodomains.py::TestTransformBiodomains::test_transform_genes_biodomains_should_pass[Pass with good data] PASSED [ 60%]
tests/transform/test_genes_biodomains.py::TestTransformBiodomains::test_transform_genes_biodomains_should_pass[Pass with imperfect data] PASSED [ 80%]
tests/transform/test_genes_biodomains.py::TestTransformBiodomains::test_transform_genes_biodomains_should_fail[Fail with bad data] PASSED [100%]

[thomasyu888]

Awesome work - I know this is WIP, but here are some initial thoughts.

[thomasyu888]

Point of discussion: Should we have these test files within transform folders:

tests/test_data/biodomains/input
tests/test_data/biodomains/output

[thomasyu888]

🪟 I like this framework, let's see what Jess/Jaclyn think!

[JessterB]

LGTM, but we should wait for Jaclyn to take a look too.

Out of curiosity what do the % values after the test output mean? Does that represent the % of the tests that have passed so far? If so, why is the first test 60% of the tests?

...PASSED [ 60%]
...PASSED [ 80%]
....PASSED [100%]

[BWMac]

Thanks @JessterB ! I believe that is what the % values mean. This looks weird here because I copied only the output lines that pertain to these new tests. The full output looks like this:

tests/transform/test_genes_biodomains.py::TestCountGroupedTotal::test_count_grouped_total_one_group PASSED                      [ 20%]
tests/transform/test_genes_biodomains.py::TestCountGroupedTotal::test_count_grouped_total_two_groups PASSED                     [ 40%]
tests/transform/test_genes_biodomains.py::TestTransformBiodomains::test_transform_biodomains_should_pass[Pass with good data] PASSED [ 60%]
tests/transform/test_genes_biodomains.py::TestTransformBiodomains::test_transform_biodomains_should_pass[Pass with imperfect data] PASSED [ 80%]
tests/transform/test_genes_biodomains.py::TestTransformBiodomains::test_transform_biodomains_should_fail[Fail with bad data] PASSED [100%]

With the tests for genes_biodomains.count_grouped_total also included.

[thomasyu888]

🔥 LGTM! I'm going to pre-approve first. Can you add to the CONTRIBUTING.md some instructions on how to contribute these transform tests? (There's a PR open for the contributing guide)

It doesn't have to be perfect, but it'll serve as a starting place for Jess/Jaclyn.

[BWMac]

@thomasyu888 I added instructions for contributing transform tests to #72. Merging this one now.