The python script runs soda core and stores results locally. The 'checks.yml' file has all the data quality checks required for each dataset.

Note that a dataset in BQ is different from a dataset in soda in that soda considers each dataset as matrix of columns and rows.