Better fix

dcherian · dcherian · commit 5cd7bcb5b497 · 2022-11-04T18:23:39.000-06:00
diff --git a/flox/core.py b/flox/core.py
@@ -1350,7 +1350,10 @@ def dask_groupby_agg(
                         aggregate=partial(aggregate, expected_groups=index, reindex=True),
                     )
                 )
-                groups_.append(cohort)
+                # This is done because pandas promotes to 64-bit types when an Index is created
+                # So we use the index to generate the return value for consistency with "map-reduce"
+                # This is important on windows
+                groups_.append(index.values)
 
             reduced = dask.array.concatenate(reduced_, axis=-1)
             groups = (np.concatenate(groups_),)

Original file line number	Diff line number	Diff line change
`@@ -1350,7 +1350,10 @@ def dask_groupby_agg(`
`1350`	`1350`	`aggregate=partial(aggregate, expected_groups=index, reindex=True),`
`1351`	`1351`	`)`
`1352`	`1352`	`)`
`1353`		`- groups_.append(cohort)`
	`1353`	`+ # This is done because pandas promotes to 64-bit types when an Index is created`
	`1354`	`+ # So we use the index to generate the return value for consistency with "map-reduce"`
	`1355`	`+ # This is important on windows`
	`1356`	`+ groups_.append(index.values)`
`1354`	`1357`
`1355`	`1358`	`reduced = dask.array.concatenate(reduced_, axis=-1)`
`1356`	`1359`	`groups = (np.concatenate(groups_),)`