fix: minor basic stats quality fixes #2521
Conversation
Codecov ReportAll modified and coverable lines are covered by tests ✅
Flags with carried forward coverage won't be shown. Click here to find out more.
... and 2 files with indirect coverage changes 🚀 New features to boost your workflow:
|
| @@ -157,7 +157,7 @@ def fit(self, data, sample_weight=None, queue=None): | |||
| data_table, weights_table = to_table(data, sample_weight, queue=queue) | |||
|
|
|||
| dtype = data_table.dtype | |||
| raw_result = raw_result = self._compute_raw( | |||
| raw_result = self._compute_raw( | |||
| @@ -48,7 +48,7 @@ def generate_data(par, size, seed=777): | |||
|
|
|||
| params_spmd = {"ns": 19, "nf": 31} | |||
|
|
|||
| data, weights = generate_data(params_spmd, size) | |||
| data, weights = generate_data(params_spmd, rank) | |||
There was a problem hiding this comment.
Shouldn't this be made to generate different data for each rank?
There was a problem hiding this comment.
Yes - that was the original mistake here. size is the same for every rank, so the same data is generated. rank is different on every rank, so different data is generated.
There was a problem hiding this comment.
But it's still being generated in a loop where each rank contains the data from the previous one.
There was a problem hiding this comment.
Although maybe that would be reflected in the seed parameter and this should be tweaked further. The data generation function here is pretty wonky. I'll take a closer look tomorrow.
* fix: minor basic stats quality fixes * blacked * vary seed by rank instead of size
Description
A few small corrections
PR completeness and readability
Testing