fix(python): fix use of date_range with 'lazy' parameter (#5652)

pola-rs · Nov 28, 2022 · 8497f95 · 8497f95
1 parent 9eaf247
commit 8497f95
Show file tree

Hide file tree

Showing 3 changed files with 74 additions and 9 deletions.
diff --git a/py-polars/polars/internals/functions.py b/py-polars/polars/internals/functions.py
@@ -298,9 +298,9 @@ def date_range(
     Using polars duration string to specify the interval:
 
     >>> from datetime import date
-    >>> pl.date_range(date(2022, 1, 1), date(2022, 3, 1), "1mo", name="drange")
+    >>> pl.date_range(date(2022, 1, 1), date(2022, 3, 1), "1mo", name="dtrange")
     shape: (3,)
-    Series: 'drange' [date]
+    Series: 'dtrange' [date]
     [
         2022-01-01
         2022-02-01
@@ -329,14 +329,16 @@ def date_range(
     ]
 
     """
+    if name is None:
+        name = ""
     if isinstance(interval, timedelta):
         interval = _timedelta_to_pl_duration(interval)
     elif " " in interval:
         interval = interval.replace(" ", "")
 
     if isinstance(low, pli.Expr) or isinstance(high, pli.Expr) or lazy:
-        low = pli.expr_to_lit_or_expr(low, str_to_lit=True)
-        high = pli.expr_to_lit_or_expr(high, str_to_lit=True)
+        low = pli.expr_to_lit_or_expr(low, str_to_lit=True)._pyexpr
+        high = pli.expr_to_lit_or_expr(high, str_to_lit=True)._pyexpr
         return pli.wrap_expr(
             _py_date_range_lazy(low, high, interval, closed, name, time_zone)
         )
@@ -370,9 +372,6 @@ def date_range(
 
     start = _datetime_to_pl_timestamp(low, tu)
     stop = _datetime_to_pl_timestamp(high, tu)
-    if name is None:
-        name = ""
-
     dt_range = pli.wrap_s(
         _py_date_range(start, stop, interval, closed, name, tu, time_zone)
     )

diff --git a/py-polars/tests/unit/test_datelike.py b/py-polars/tests/unit/test_datelike.py
@@ -516,6 +516,72 @@ def test_date_range() -> None:
     assert result.cast(pl.Utf8)[-1] == "2022-01-01 00:00:59.247379260"
 
 
+def test_date_range_lazy() -> None:
+    # lazy date range with literals
+    df = pl.DataFrame({"misc": ["x"]}).with_columns(
+        pl.date_range(
+            date(2000, 1, 1),
+            date(2023, 8, 31),
+            interval="987d",
+            lazy=True,
+        )
+        .list()
+        .alias("dts")
+    )
+    assert df.rows() == [
+        (
+            "x",
+            [
+                date(2000, 1, 1),
+                date(2002, 9, 14),
+                date(2005, 5, 28),
+                date(2008, 2, 9),
+                date(2010, 10, 23),
+                date(2013, 7, 6),
+                date(2016, 3, 19),
+                date(2018, 12, 1),
+                date(2021, 8, 14),
+            ],
+        )
+    ]
+    assert (
+        df.rows()[0][1]
+        == pd.date_range(
+            date(2000, 1, 1), date(2023, 12, 31), freq="987d"
+        ).date.tolist()
+    )
+
+    # lazy date range with expressions
+    ldf = (
+        pl.DataFrame({"start": [date(2015, 6, 30)], "stop": [date(2022, 12, 31)]})
+        .with_columns(
+            pl.date_range(
+                pl.col("start"),
+                pl.col("stop"),
+                interval="678d",
+                lazy=True,
+            )
+            .list()
+            .alias("dts")
+        )
+        .lazy()
+    )
+
+    assert ldf.collect().rows() == [
+        (
+            date(2015, 6, 30),
+            date(2022, 12, 31),
+            [
+                date(2015, 6, 30),
+                date(2017, 5, 8),
+                date(2019, 3, 17),
+                date(2021, 1, 23),
+                date(2022, 12, 2),
+            ],
+        )
+    ]
+
+
 @pytest.mark.parametrize(
     "one,two",
     [

diff --git a/py-polars/tests/unit/test_df.py b/py-polars/tests/unit/test_df.py
@@ -1387,10 +1387,10 @@ def test_reproducible_hash_with_seeds() -> None:
     seeds = (11, 22, 33, 44)
 
     # TODO: introduce a platform-stable string hash...
-    #  in the meantime, account for arm64 (mac) hash values to reduce noise
+    #  in the meantime, try to account for arm64 (mac) hash values to reduce noise
     expected = pl.Series(
         "s",
-        [8823051245921001677, 988796329533502010, 7528667241828618484]
+        [6629530352159708028,15496313222292466864,6048298245521876612]
         if platform.mac_ver()[-1] == "arm64"
         else [6629530352159708028, 988796329533502010, 6048298245521876612],
         dtype=pl.UInt64,