Validation aggregation factory

[tylerriccio33]

Summary

This is a PR to create aggregation methods for the Validate class. Users can compare summary statistics like sum, sd, avg, etc. I'm trying to make this as modular as possible and dynamically generate the methods themselves.

This is a WIP so no need to merge. I'll have to at least implement the most common stats.

Related GitHub Issues and PRs

This PR references the conversation about implementing higher order methods and wrappers specifically for summary stats.

• Ref: Add percentage null assertion #290

[tylerriccio33]

Alright so I implemented a bare bonescol_sum_* and col_avg_*. Do you have any immediate concerns or pointers? If not I'll go ahead and implement more aggregators.

Also, why is this test failing "FAILED tests/test_validate.py::test_validation_with_all_validation_types_pandas - assert (18 / 21) >= 0.9" ? It failed in the other local-tests branch you merged in where there were no real changes, so makes me think it's spurious. Any ideas?

[rich-iannone]

The .col_sum_*() and .col_avg_*() methods look great! No concerns at all from looking at the new files (I'll come up with the set of icons soon, as before).

And I think it's a recent release of Narwhals that broke the Pandas test. I'll fix that and once that's merged, you can merge this PR to main and it should no longer be a problem.

[MarcoGorelli]

And I think it's a recent release of Narwhals that broke the Pandas test. I'll fix that and once that's merged, you can merge this PR to main and it should no longer be a problem.

sorry about that, though i'm a little confused as to how, given that we're meant to be running pointblack tests in our ci https://github.com/narwhals-dev/narwhals/actions/runs/19853532931/job/56885727968 -taking a look 👀

---

this passes for me locally 🤷 any idea what the issue might be?

[rich-iannone]

@MarcoGorelli looking back into the CI logs, I actually can't find any trace of the failure in the merged PR's commits. And it never happened for me in local test runs.

I only suspected it might have something to do with Narwhals because of the recent update around that time (but I should have remembered to check the Narwhals CI for the downstream libraries testing). That said, the failure on CI was fixed with only a small test adjustment.