More ruff updates

Author: timsaucer

python/tests/generic.py

@@ -16,6 +16,7 @@
 # under the License.
 
 import datetime
+from datetime import timezone
 
 import numpy as np
 import pyarrow as pa
@@ -26,29 +27,29 @@
 
 
 def data():
-    np.random.seed(1)
+    rng = np.random.default_rng(1)
     data = np.concatenate(
         [
-            np.random.normal(0, 0.01, size=50),
-            np.random.normal(50, 0.01, size=50),
+            rng.normal(0, 0.01, size=50),
+            rng.normal(50, 0.01, size=50),
         ]
     )
     return pa.array(data)
 
 
 def data_with_nans():
-    np.random.seed(0)
-    data = np.random.normal(0, 0.01, size=50)
-    mask = np.random.randint(0, 2, size=50)
+    rng = np.random.default_rng(0)
+    data = rng.normal(0, 0.01, size=50)
+    mask = rng.normal(0, 2, size=50)
     data[mask == 0] = np.nan
     return data
 
 
 def data_datetime(f):
     data = [
-        datetime.datetime.now(),
-        datetime.datetime.now() - datetime.timedelta(days=1),
-        datetime.datetime.now() + datetime.timedelta(days=1),
+        datetime.datetime.now(tz=timezone.utc),
+        datetime.datetime.now(tz=timezone.utc) - datetime.timedelta(days=1),
+        datetime.datetime.now(tz=timezone.utc) + datetime.timedelta(days=1),
     ]
     return pa.array(data, type=pa.timestamp(f), mask=np.array([False, True, False]))
 

python/tests/test_aggregation.py

@@ -66,7 +66,7 @@ def df_aggregate_100():
 
 
 @pytest.mark.parametrize(
-    "agg_expr, calc_expected",
+    ("agg_expr", "calc_expected"),
     [
         (f.avg(column("a")), lambda a, b, c, d: np.array(np.average(a))),
         (
@@ -88,7 +88,7 @@ def df_aggregate_100():
             f.covar_samp(column("b"), column("c")),
             lambda a, b, c, d: np.array(np.cov(b, c, ddof=1)[0][1]),
         ),
-        # f.grouping(col_a),  # No physical plan implemented yet
+        # f.grouping(col_a),  # No physical plan implemented yet  # noqa: ERA001
         (f.max(column("a")), lambda a, b, c, d: np.array(np.max(a))),
         (f.mean(column("b")), lambda a, b, c, d: np.array(np.mean(b))),
         (f.median(column("b")), lambda a, b, c, d: np.array(np.median(b))),
@@ -114,7 +114,7 @@ def test_aggregation_stats(df, agg_expr, calc_expected):
 
 
 @pytest.mark.parametrize(
-    "agg_expr, expected, array_sort",
+    ("agg_expr", "expected", "array_sort"),
     [
         (f.approx_distinct(column("b")), pa.array([2], type=pa.uint64()), False),
         (
@@ -182,12 +182,11 @@ def test_aggregation(df, agg_expr, expected, array_sort):
     agg_df.show()
     result = agg_df.collect()[0]
 
-    print(result)
     assert result.column(0) == expected
 
 
 @pytest.mark.parametrize(
-    "name,expr,expected",
+    ("name", "expr", "expected"),
     [
         (
             "approx_percentile_cont",
@@ -313,7 +312,7 @@ def test_bit_and_bool_fns(df, name, expr, result):
 
 
 @pytest.mark.parametrize(
-    "name,expr,result",
+    ("name", "expr", "result"),
     [
         ("first_value", f.first_value(column("a")), [0, 4]),
         (
@@ -363,7 +362,6 @@ def test_bit_and_bool_fns(df, name, expr, result):
             ),
             [8, 9],
         ),
-        ("first_value", f.first_value(column("a")), [0, 4]),
         (
             "nth_value_ordered",
             f.nth_value(column("a"), 2, order_by=[column("a").sort(ascending=False)]),
@@ -403,7 +401,7 @@ def test_first_last_value(df_partitioned, name, expr, result) -> None:
 
 
 @pytest.mark.parametrize(
-    "name,expr,result",
+    ("name", "expr", "result"),
     [
         ("string_agg", f.string_agg(column("a"), ","), "one,two,three,two"),
         ("string_agg", f.string_agg(column("b"), ""), "03124"),

python/tests/test_catalog.py

@@ -19,17 +19,20 @@
 import pytest
 
 
+# Note we take in `database` as a variable even though we don't use
+# it because that will cause the fixture to set up the context with
+# the tables we need.
 def test_basic(ctx, database):
     with pytest.raises(KeyError):
         ctx.catalog("non-existent")
 
     default = ctx.catalog()
     assert default.names() == ["public"]
 
-    for database in [default.database("public"), default.database()]:
-        assert database.names() == {"csv1", "csv", "csv2"}
+    for db in [default.database("public"), default.database()]:
+        assert db.names() == {"csv1", "csv", "csv2"}
 
-    table = database.table("csv")
+    table = db.table("csv")
     assert table.kind == "physical"
     assert table.schema == pa.schema(
         [