darkreactions / ESCALATE_report

Transform experimental data into ML ready datasets!
http://escalation.sd2e.org/
MIT License
2 stars 1 forks source link

Output validation #30

Closed miketynes closed 5 years ago

miketynes commented 5 years ago

A quick-and-dirty output validator.

Run this after a change to the codebase to ensure debug output hasn't changed. Intended usage usage:

outputvalidation.py will ignore new columns in the target_output (that do not appear in reference_output) CSV automatically. If you expect any columns to be different across the two output files, specify them after the --ignore flag.

Currently outputvalidation.py does not support intentionally removing variables from the report output, but this could be added easily. That is to say if a variable appears in reference_output.csv but not target_output.csv, this will yield NARP

Testing Note: if you intend to test the functionality of this code by modifying output CSVs by hand in Excel, make sure you open and save both files from Excel. Simply opening one of them and pressing save will change the way the data is formatted, causing outputvalidation.py to return NARP