Closed VDFaller closed 8 months ago
Great catch, thanks for the PR! I think this is one of those holdovers from Great Expectations where tests are built against static values vs totals of other columns. Probably worth looking other tests as well.
Would you mind also adding a test that covers this new use case of testing against the sum of another column? We currently only have a static test, i.e.
- dbt_expectations.expect_multicolumn_sum_to_equal:
column_list: ["col_numeric_a", "col_numeric_b"]
sum_total: 4
so we'll need something like
- dbt_expectations.expect_multicolumn_sum_to_equal:
column_list: ["col_numeric_a", "col_numeric_b"]
sum_total: new_totals_column
Hopefully this is what you're looking for. If not let me know.
Hopefully this is what you're looking for. If not let me know.
I'll check it out! FYI, you probably figured this out since, but sum(4)
against the table return 16
b/c it does this for every one of the 4 rows. This may actually mean this is a breaking change. Any config that checks for 4 would now fail. Any thoughts?
FYI, updated the CI config on this PR to get Spark CI to pass. For some reason our CI configs breaks in dbt 1.7.x, but I'll fix that separately. CI is passing for you now 👍
Issue this PR Addresses/Closes
Closes #290
Summary of Changes
Just wrapped
sum_total
in Sum()Why Do We Need These Changes
This will allow a column name or a literal number to be used.
Reviewers
@clausherther