Coverage tool

[daisieh]

To test, use the data from https://github.com/CanDIG/mohccn-data. Check out this branch in the clinical_ETL_code submodule there, and run:

python validate_coverage.py --input ../Synthetic_Clinical_Data_2 --manifest ../mappings/synthetic_clinical/manifest.yml

The tool will:

• validate the mapping with the katsu internal schema and report any mismatches.
• quantify how much of the provided clinical data is covered by the mapping.
• report which schema fields are not used by the mapping at all.

[yavyx]

• Coverage tool works as expected with synthetic data.
• Add instructions to use coverage tool on README.
• Consider using a DataFrame to store the mapping coverage table for pretty printing and easy cross-cohort comparisons/analyses.
• Consider adding docstrings to functions to briefly describe their inputs and outputs.

BC POG data throws an error but appears to work properly anyways.

[kcranston]

This is awesome! I agree with @yavyx that using a dataframe for a pretty printing option would be nice (since we already have a pandas requirement). That can be its own issue, though. Doesn't need to hold this up.

Yay! Data reports!