`cnv_discordant_read_calls` raises an `IndexError` instead of a `ValueError` for absent data

[jonbrenas]

When trying to access the discordant read calls for af1 (which do not exist, I gather), I encountered an unexpected error:

It turns out the code contains a test handling the absence of data for a contig/set of samples but the check is indented too much (in my opinion):

[Jiya873]

@jonbrenas, I have raised a pull request #781 addressing the recent changes. Could you please take a moment to review it?

[AswaniSahoo]

Hi @jonbrenas! I see that the previous PR (#781) for this issue appears to have been abandoned since August 2025, with the tests still failing due to monkeypatching issues and merge conflicts with master.

I'd like to take a fresh approach to resolve this. My plan:

1. Add an explicit empty-data check in cnv_discordant_read_calls before any indexing occurs raising a descriptive ValueError that tells the user which sample set/contig combination had no data
2. Write tests properly using the existing simulated data fixtures (avoiding monkeypatching, which caused the previous PR's test failures), following the same patterns used in other test files like test_snp_data.py
3. Ensure full test coverage the previous PR had only 50% patch coverage

I have experience with similar error handling patterns in scientific data APIs my biodiversity-publication-analyzer handles absent-data edge cases with informative messages across genomic data pipelines using Europe PMC.

I'll base my work off current master to avoid the merge conflict issues. Happy to open a PR once I have a clean implementation.

[jonbrenas]

Hi @AswaniSahoo. Feel free to create a PR. I think there is a very simple solution to this issue but any code that would make things clearer is welcome.