Change in metadata structure of Pf9

[jonbrenas]

@nw20, @podpearson, it looks like there is an error in the integration tests due to some change in the content of the metadata for Pf9.

FAILED tests/integration/test_pf9.py::test_sample_metadata[gs://pf9-release/] - AssertionError: assert ('Sample', 'S...ocation', ...) == ('Sample', 'S...ocation', ...)

At index 17 diff: 'Curated timepoint' != 'Source ENA accession'

Left contains 2 more items, first extra item: 'MalariaGEN ENA CRAM URL'

Full diff:

(
'Sample',
'Study',
'Country',
'Admin level 1',
'Admin level 2',
'Curated location',
'Country latitude',
'Country longitude',
'Admin level 1 latitude',
'Admin level 1 longitude',
'Admin level 2 latitude',
'Admin level 2 longitude',
'Location latitude',
'Location longitude',
'Year',
'Month',
'Day',
+     'Curated timepoint',
'Source ENA accession',
'Alternative ID',
'All samples same case',
'Population',
'% callable',
'QC pass',
+     'QC pass plus time series',
'Exclusion reason',
'Sample type',
'Sample was in Pf8',
'MalariaGEN ENA project accession',
'MalariaGEN ENA experiment accession',
'MalariaGEN ENA sample accession',
'MalariaGEN ENA run accession',
'MalariaGEN ENA CRAM URL',
'MalariaGEN ENA FASTQ URL',
)

I assume the solution is to update the test to use the correct expected set of columns (or edit the columns to remove the 2 new ones but that sounds wrong).

[podpearson]

@jonbrenas - yes, I think updating the test to use the correct set of columns should be the right solution. Thanks!

[jonbrenas]

Thanks, @podpearson.

[jonbrenas]

@podpearson, @nw20, the best solution would probably be to remove the integration tests from the public repo while Pf9 is not officially public. They can be reinstated when we start granting people access to the resource. Would you mind doing that?