fix: fix hashing specialization (#14754)

pola-rs · Feb 29, 2024 · 8d0cdc2 · 8d0cdc2
1 parent a2f5cf2
commit 8d0cdc2
Show file tree

Hide file tree

Showing 2 changed files with 18 additions and 2 deletions.
diff --git a/crates/polars-core/src/frame/group_by/hashing.rs b/crates/polars-core/src/frame/group_by/hashing.rs
@@ -223,7 +223,7 @@ where
                                 RawEntryMut::Vacant(entry) => {
                                     let tuples = unitvec![idx];
                                     entry.insert_with_hasher(hash, k, (idx, tuples), |k| {
-                                        hasher.hash_one(k)
+                                        hasher.hash_one(*k)
                                     });
                                 },
                                 RawEntryMut::Occupied(mut entry) => {
@@ -288,7 +288,7 @@ where
                                 RawEntryMut::Vacant(entry) => {
                                     let tuples = unitvec![idx];
                                     entry.insert_with_hasher(hash, k, (idx, tuples), |k| {
-                                        hasher.hash_one(k)
+                                        hasher.hash_one(*k)
                                     });
                                 },
                                 RawEntryMut::Occupied(mut entry) => {

diff --git a/py-polars/tests/unit/operations/test_aggregations.py b/py-polars/tests/unit/operations/test_aggregations.py
@@ -423,3 +423,19 @@ def test_agg_empty_sum_after_filter_14734() -> None:
     expect = pl.Series("b", [0, 0]).to_frame()
     assert_frame_equal(expect, last.select("b"))
     assert_frame_equal(expect, curr.select("b"))
+
+
+@pytest.mark.slow()
+def test_grouping_hash_14749() -> None:
+    n_groups = 251
+    rows_per_group = 4
+    assert (
+        pl.DataFrame(
+            {
+                "grp": np.repeat(np.arange(n_groups), rows_per_group),
+                "x": np.tile(np.arange(rows_per_group), n_groups),
+            }
+        )
+        .select(pl.col("x").max().over("grp"))["x"]
+        .value_counts()
+    ).to_dict(as_series=False) == {"x": [3], "count": [1004]}