From 82da0892732e4c1058d769c93cc3b6a424282cde Mon Sep 17 00:00:00 2001
From: Francesco Bruzzesi <42817048+FBruzzesi@users.noreply.github.com>
Date: Mon, 20 Jan 2025 07:45:51 +0100
Subject: [PATCH] test: decouple pyspark constructor from pandas (#1834)

---
 tests/conftest.py                             | 42 +++++++++-------
 tests/expr_and_series/arg_max_test.py         |  2 +-
 tests/expr_and_series/arg_min_test.py         |  2 +-
 tests/expr_and_series/arithmetic_test.py      | 16 +++----
 tests/expr_and_series/binary_test.py          |  2 +-
 tests/expr_and_series/double_selected_test.py |  2 +-
 tests/expr_and_series/double_test.py          |  4 +-
 tests/expr_and_series/fill_null_test.py       |  8 ++--
 tests/expr_and_series/is_duplicated_test.py   |  4 +-
 tests/expr_and_series/is_unique_test.py       |  4 +-
 tests/expr_and_series/lit_test.py             |  4 +-
 tests/expr_and_series/max_test.py             |  2 +-
 tests/expr_and_series/mean_test.py            |  2 +-
 tests/expr_and_series/median_test.py          |  2 +-
 tests/expr_and_series/min_test.py             |  2 +-
 tests/expr_and_series/n_unique_test.py        |  2 +-
 tests/expr_and_series/nth_test.py             |  4 +-
 tests/expr_and_series/null_count_test.py      |  2 +-
 tests/expr_and_series/quantile_test.py        |  4 +-
 tests/expr_and_series/std_test.py             |  4 +-
 tests/expr_and_series/sum_horizontal_test.py  |  2 +-
 tests/expr_and_series/sum_test.py             |  2 +-
 tests/expr_and_series/unary_test.py           |  4 +-
 tests/expr_and_series/var_test.py             |  4 +-
 tests/frame/add_test.py                       |  2 +-
 tests/frame/columns_test.py                   |  2 +-
 tests/frame/concat_test.py                    |  6 +--
 tests/frame/double_test.py                    |  2 +-
 tests/frame/drop_test.py                      |  2 +-
 tests/frame/filter_test.py                    |  8 ++--
 tests/frame/head_test.py                      |  2 +-
 tests/frame/invalid_test.py                   |  2 +-
 tests/frame/is_duplicated_test.py             |  2 +-
 tests/frame/is_empty_test.py                  |  2 +-
 tests/frame/is_unique_test.py                 |  2 +-
 tests/frame/item_test.py                      |  4 +-
 tests/frame/join_test.py                      | 48 +++++++++----------
 tests/frame/reindex_test.py                   |  2 +-
 tests/frame/rename_test.py                    |  4 +-
 tests/frame/rows_test.py                      |  6 +--
 tests/frame/schema_test.py                    |  8 +++-
 tests/frame/select_test.py                    |  4 +-
 tests/frame/sort_test.py                      |  2 +-
 tests/frame/tail_test.py                      |  2 +-
 tests/frame/to_arrow_test.py                  |  2 +-
 tests/frame/to_dict_test.py                   |  4 +-
 tests/frame/to_native_test.py                 |  2 +-
 tests/frame/to_numpy_test.py                  |  4 +-
 tests/frame/to_pandas_test.py                 |  2 +-
 tests/frame/to_polars_test.py                 |  2 +-
 tests/frame/unique_test.py                    |  2 +-
 tests/frame/with_columns_test.py              |  8 ++--
 tests/group_by_test.py                        |  2 +-
 tests/series_only/is_sorted_test.py           |  2 +-
 54 files changed, 140 insertions(+), 128 deletions(-)

diff --git a/tests/conftest.py b/tests/conftest.py
index 95b969e95f..eeb1f1a820 100644
--- a/tests/conftest.py
+++ b/tests/conftest.py
@@ -2,6 +2,7 @@
 
 import os
 import sys
+from copy import deepcopy
 from typing import TYPE_CHECKING
 from typing import Any
 from typing import Callable
@@ -12,6 +13,8 @@
 import pyarrow as pa
 import pytest
 
+from narwhals.utils import generate_temporary_column_name
+
 if TYPE_CHECKING:
     import duckdb
 
@@ -67,64 +70,66 @@ def pytest_collection_modifyitems(
             item.add_marker(skip_slow)
 
 
-def pandas_constructor(obj: Any) -> IntoDataFrame:
+def pandas_constructor(obj: dict[str, list[Any]]) -> IntoDataFrame:
     return pd.DataFrame(obj)  # type: ignore[no-any-return]
 
 
-def pandas_nullable_constructor(obj: Any) -> IntoDataFrame:
+def pandas_nullable_constructor(obj: dict[str, list[Any]]) -> IntoDataFrame:
     return pd.DataFrame(obj).convert_dtypes(dtype_backend="numpy_nullable")  # type: ignore[no-any-return]
 
 
-def pandas_pyarrow_constructor(obj: Any) -> IntoDataFrame:
+def pandas_pyarrow_constructor(obj: dict[str, list[Any]]) -> IntoDataFrame:
     return pd.DataFrame(obj).convert_dtypes(dtype_backend="pyarrow")  # type: ignore[no-any-return]
 
 
-def modin_constructor(obj: Any) -> IntoDataFrame:  # pragma: no cover
+def modin_constructor(obj: dict[str, list[Any]]) -> IntoDataFrame:  # pragma: no cover
     import modin.pandas as mpd
 
     return mpd.DataFrame(pd.DataFrame(obj))  # type: ignore[no-any-return]
 
 
-def modin_pyarrow_constructor(obj: Any) -> IntoDataFrame:  # pragma: no cover
+def modin_pyarrow_constructor(
+    obj: dict[str, list[Any]],
+) -> IntoDataFrame:  # pragma: no cover
     import modin.pandas as mpd
 
     return mpd.DataFrame(pd.DataFrame(obj)).convert_dtypes(dtype_backend="pyarrow")  # type: ignore[no-any-return]
 
 
-def cudf_constructor(obj: Any) -> IntoDataFrame:  # pragma: no cover
+def cudf_constructor(obj: dict[str, list[Any]]) -> IntoDataFrame:  # pragma: no cover
     import cudf
 
     return cudf.DataFrame(obj)  # type: ignore[no-any-return]
 
 
-def polars_eager_constructor(obj: Any) -> IntoDataFrame:
+def polars_eager_constructor(obj: dict[str, list[Any]]) -> IntoDataFrame:
     return pl.DataFrame(obj)
 
 
-def polars_lazy_constructor(obj: Any) -> pl.LazyFrame:
+def polars_lazy_constructor(obj: dict[str, list[Any]]) -> pl.LazyFrame:
     return pl.LazyFrame(obj)
 
 
-def duckdb_lazy_constructor(obj: Any) -> duckdb.DuckDBPyRelation:
+def duckdb_lazy_constructor(obj: dict[str, list[Any]]) -> duckdb.DuckDBPyRelation:
     import duckdb
 
     _df = pl.LazyFrame(obj)
     return duckdb.table("_df")
 
 
-def dask_lazy_p1_constructor(obj: Any) -> IntoFrame:  # pragma: no cover
+def dask_lazy_p1_constructor(obj: dict[str, list[Any]]) -> IntoFrame:  # pragma: no cover
     import dask.dataframe as dd
 
     return dd.from_dict(obj, npartitions=1)  # type: ignore[no-any-return]
 
 
-def dask_lazy_p2_constructor(obj: Any) -> IntoFrame:  # pragma: no cover
+def dask_lazy_p2_constructor(obj: dict[str, list[Any]]) -> IntoFrame:  # pragma: no cover
     import dask.dataframe as dd
 
     return dd.from_dict(obj, npartitions=2)  # type: ignore[no-any-return]
 
 
-def pyarrow_table_constructor(obj: Any) -> IntoDataFrame:
+def pyarrow_table_constructor(obj: dict[str, list[Any]]) -> IntoDataFrame:
     return pa.table(obj)  # type: ignore[no-any-return]
 
 
@@ -159,13 +164,16 @@ def pyspark_lazy_constructor() -> Callable[[Any], IntoFrame]:  # pragma: no cove
 
         register(session.stop)
 
-        def _constructor(obj: Any) -> IntoFrame:
-            pd_df = pd.DataFrame(obj).replace({float("nan"): None}).reset_index()
+        def _constructor(obj: dict[str, list[Any]]) -> IntoFrame:
+            _obj = deepcopy(obj)
+            index_col_name = generate_temporary_column_name(n_bytes=8, columns=list(_obj))
+            _obj[index_col_name] = list(range(len(_obj[next(iter(_obj))])))
+
             return (  # type: ignore[no-any-return]
-                session.createDataFrame(pd_df)
+                session.createDataFrame([*zip(*_obj.values())], schema=[*_obj.keys()])
                 .repartition(2)
-                .orderBy("index")
-                .drop("index")
+                .orderBy(index_col_name)
+                .drop(index_col_name)
             )
 
         return _constructor
diff --git a/tests/expr_and_series/arg_max_test.py b/tests/expr_and_series/arg_max_test.py
index 45b1ad0610..9f9de07747 100644
--- a/tests/expr_and_series/arg_max_test.py
+++ b/tests/expr_and_series/arg_max_test.py
@@ -6,7 +6,7 @@
 from tests.utils import ConstructorEager
 from tests.utils import assert_equal_data
 
-data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8, 9], "i": [3, 1, 5]}
+data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0], "i": [3, 1, 5]}
 
 
 def test_expr_arg_max_expr(
diff --git a/tests/expr_and_series/arg_min_test.py b/tests/expr_and_series/arg_min_test.py
index 1654f6fee5..406e80c207 100644
--- a/tests/expr_and_series/arg_min_test.py
+++ b/tests/expr_and_series/arg_min_test.py
@@ -6,7 +6,7 @@
 from tests.utils import ConstructorEager
 from tests.utils import assert_equal_data
 
-data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8, 9]}
+data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0]}
 
 
 def test_expr_arg_min_expr(
diff --git a/tests/expr_and_series/arithmetic_test.py b/tests/expr_and_series/arithmetic_test.py
index 1baae44e53..3a5cde3a36 100644
--- a/tests/expr_and_series/arithmetic_test.py
+++ b/tests/expr_and_series/arithmetic_test.py
@@ -45,7 +45,7 @@ def test_arithmetic_expr(
     ):
         request.applymarker(pytest.mark.xfail)
 
-    data = {"a": [1.0, 2, 3]}
+    data = {"a": [1.0, 2.0, 3.0]}
     df = nw.from_native(constructor(data))
     result = df.select(getattr(nw.col("a"), attr)(rhs))
     assert_equal_data(result, {"a": expected})
@@ -57,7 +57,7 @@ def test_arithmetic_expr(
         ("__radd__", 1, [2, 3, 4]),
         ("__rsub__", 1, [0, -1, -2]),
         ("__rmul__", 2, [2, 4, 6]),
-        ("__rtruediv__", 2.0, [2, 1, 2 / 3]),
+        ("__rtruediv__", 2.0, [2.0, 1.0, 2 / 3]),
         ("__rfloordiv__", 2, [2, 1, 0]),
         ("__rmod__", 2, [0, 0, 2]),
         ("__rpow__", 2, [2, 4, 8]),
@@ -119,7 +119,7 @@ def test_arithmetic_series(
         ("__radd__", 1, [2, 3, 4]),
         ("__rsub__", 1, [0, -1, -2]),
         ("__rmul__", 2, [2, 4, 6]),
-        ("__rtruediv__", 2.0, [2, 1, 2 / 3]),
+        ("__rtruediv__", 2.0, [2.0, 1.0, 2 / 3]),
         ("__rfloordiv__", 2, [2, 1, 0]),
         ("__rmod__", 2, [0, 0, 2]),
         ("__rpow__", 2, [2, 4, 8]),
@@ -231,7 +231,7 @@ def test_mod(left: int, right: int) -> None:
         ("__add__", nw.lit(1), [2, 3, 5]),
         ("__sub__", nw.lit(1), [0, -1, -3]),
         ("__mul__", nw.lit(2), [2, 4, 8]),
-        ("__truediv__", nw.lit(2.0), [2, 1, 0.5]),
+        ("__truediv__", nw.lit(2.0), [2.0, 1.0, 0.5]),
         ("__truediv__", nw.lit(1), [1, 0.5, 0.25]),
         ("__floordiv__", nw.lit(2), [2, 1, 0]),
         ("__mod__", nw.lit(3), [0, 1, 3]),
@@ -254,7 +254,7 @@ def test_arithmetic_expr_left_literal(
     ):
         request.applymarker(pytest.mark.xfail)
 
-    data = {"a": [1.0, 2, 4]}
+    data = {"a": [1.0, 2.0, 4.0]}
     df = nw.from_native(constructor(data))
     result = df.select(getattr(lhs, attr)(nw.col("a")))
     assert_equal_data(result, {"literal": expected})
@@ -266,8 +266,8 @@ def test_arithmetic_expr_left_literal(
         ("__add__", nw.lit(1), [2, 3, 5]),
         ("__sub__", nw.lit(1), [0, -1, -3]),
         ("__mul__", nw.lit(2), [2, 4, 8]),
-        ("__truediv__", nw.lit(2.0), [2, 1, 0.5]),
-        ("__truediv__", nw.lit(1), [1, 0.5, 0.25]),
+        ("__truediv__", nw.lit(2.0), [2.0, 1.0, 0.5]),
+        ("__truediv__", nw.lit(1), [1.0, 0.5, 0.25]),
         ("__floordiv__", nw.lit(2), [2, 1, 0]),
         ("__mod__", nw.lit(3), [0, 1, 3]),
         ("__pow__", nw.lit(2), [2, 4, 16]),
@@ -285,7 +285,7 @@ def test_arithmetic_series_left_literal(
     ):
         request.applymarker(pytest.mark.xfail)
 
-    data = {"a": [1.0, 2, 4]}
+    data = {"a": [1.0, 2.0, 4.0]}
     df = nw.from_native(constructor_eager(data))
     result = df.select(getattr(lhs, attr)(nw.col("a")))
     assert_equal_data(result, {"literal": expected})
diff --git a/tests/expr_and_series/binary_test.py b/tests/expr_and_series/binary_test.py
index 308745cb43..6f8e8670d2 100644
--- a/tests/expr_and_series/binary_test.py
+++ b/tests/expr_and_series/binary_test.py
@@ -13,7 +13,7 @@ def test_expr_binary(constructor: Constructor, request: pytest.FixtureRequest) -
         constructor
     ):
         request.applymarker(pytest.mark.xfail)
-    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8, 9]}
+    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0]}
     df_raw = constructor(data)
     result = nw.from_native(df_raw).with_columns(
         a=(1 + 3 * nw.col("a")) * (1 / nw.col("a")),
diff --git a/tests/expr_and_series/double_selected_test.py b/tests/expr_and_series/double_selected_test.py
index 9eb918924b..a99c901630 100644
--- a/tests/expr_and_series/double_selected_test.py
+++ b/tests/expr_and_series/double_selected_test.py
@@ -6,7 +6,7 @@
 
 
 def test_double_selected(constructor: Constructor) -> None:
-    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8, 9]}
+    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7, 8, 9]}
     df = nw.from_native(constructor(data))
 
     result = df.select(nw.col("a", "b") * 2)
diff --git a/tests/expr_and_series/double_test.py b/tests/expr_and_series/double_test.py
index 6f43ad1393..df988ab181 100644
--- a/tests/expr_and_series/double_test.py
+++ b/tests/expr_and_series/double_test.py
@@ -6,7 +6,7 @@
 
 
 def test_double(constructor: Constructor) -> None:
-    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8, 9]}
+    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0]}
     df = nw.from_native(constructor(data))
     result = df.with_columns(nw.all() * 2)
     expected = {"a": [2, 6, 4], "b": [8, 8, 12], "z": [14.0, 16.0, 18.0]}
@@ -14,7 +14,7 @@ def test_double(constructor: Constructor) -> None:
 
 
 def test_double_alias(constructor: Constructor) -> None:
-    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8, 9]}
+    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0]}
     df = nw.from_native(constructor(data))
     result = df.with_columns(nw.col("a").alias("o"), nw.all() * 2)
     expected = {
diff --git a/tests/expr_and_series/fill_null_test.py b/tests/expr_and_series/fill_null_test.py
index 39b0a3c648..97f93768f8 100644
--- a/tests/expr_and_series/fill_null_test.py
+++ b/tests/expr_and_series/fill_null_test.py
@@ -16,9 +16,9 @@ def test_fill_null(request: pytest.FixtureRequest, constructor: Constructor) ->
     if "pyspark" in str(constructor):
         request.applymarker(pytest.mark.xfail)
     data = {
-        "a": [0.0, None, 2, 3, 4],
-        "b": [1.0, None, None, 5, 3],
-        "c": [5.0, None, 3, 2, 1],
+        "a": [0.0, None, 2.0, 3.0, 4.0],
+        "b": [1.0, None, None, 5.0, 3.0],
+        "c": [5.0, None, 3.0, 2.0, 1.0],
     }
     df = nw.from_native(constructor(data))
 
@@ -33,7 +33,7 @@ def test_fill_null(request: pytest.FixtureRequest, constructor: Constructor) ->
 
 def test_fill_null_exceptions(constructor: Constructor) -> None:
     data = {
-        "a": [0.0, None, 2, 3, 4],
+        "a": [0.0, None, 2.0, 3.0, 4.0],
     }
     df = nw.from_native(constructor(data))
 
diff --git a/tests/expr_and_series/is_duplicated_test.py b/tests/expr_and_series/is_duplicated_test.py
index d97d30cbd6..fe8b45bf15 100644
--- a/tests/expr_and_series/is_duplicated_test.py
+++ b/tests/expr_and_series/is_duplicated_test.py
@@ -11,7 +11,7 @@
 def test_is_duplicated_expr(
     constructor: Constructor, request: pytest.FixtureRequest
 ) -> None:
-    if ("pyspark" in str(constructor)) or "duckdb" in str(constructor):
+    if "duckdb" in str(constructor):
         request.applymarker(pytest.mark.xfail)
     data = {"a": [1, 1, 2], "b": [1, 2, 3], "index": [0, 1, 2]}
     df = nw.from_native(constructor(data))
@@ -23,7 +23,7 @@ def test_is_duplicated_expr(
 def test_is_duplicated_w_nulls_expr(
     constructor: Constructor, request: pytest.FixtureRequest
 ) -> None:
-    if ("pyspark" in str(constructor)) or "duckdb" in str(constructor):
+    if "duckdb" in str(constructor):
         request.applymarker(pytest.mark.xfail)
     data = {"a": [1, 1, None], "b": [1, None, None], "index": [0, 1, 2]}
     df = nw.from_native(constructor(data))
diff --git a/tests/expr_and_series/is_unique_test.py b/tests/expr_and_series/is_unique_test.py
index 92e7256234..3e9259c03a 100644
--- a/tests/expr_and_series/is_unique_test.py
+++ b/tests/expr_and_series/is_unique_test.py
@@ -9,7 +9,7 @@
 
 
 def test_is_unique_expr(constructor: Constructor, request: pytest.FixtureRequest) -> None:
-    if ("pyspark" in str(constructor)) or "duckdb" in str(constructor):
+    if "duckdb" in str(constructor):
         request.applymarker(pytest.mark.xfail)
     data = {
         "a": [1, 1, 2],
@@ -29,7 +29,7 @@ def test_is_unique_expr(constructor: Constructor, request: pytest.FixtureRequest
 def test_is_unique_w_nulls_expr(
     constructor: Constructor, request: pytest.FixtureRequest
 ) -> None:
-    if ("pyspark" in str(constructor)) or "duckdb" in str(constructor):
+    if "duckdb" in str(constructor):
         request.applymarker(pytest.mark.xfail)
     data = {
         "a": [None, 1, 2],
diff --git a/tests/expr_and_series/lit_test.py b/tests/expr_and_series/lit_test.py
index f24e6d4a12..b29ab89ee0 100644
--- a/tests/expr_and_series/lit_test.py
+++ b/tests/expr_and_series/lit_test.py
@@ -29,7 +29,7 @@ def test_lit(
 ) -> None:
     if "pyspark" in str(constructor) and dtype is not None:
         request.applymarker(pytest.mark.xfail)
-    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8, 9]}
+    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0]}
     df_raw = constructor(data)
     df = nw.from_native(df_raw).lazy()
     result = df.with_columns(nw.lit(2, dtype).alias("lit"))
@@ -43,7 +43,7 @@ def test_lit(
 
 
 def test_lit_error(constructor: Constructor) -> None:
-    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8, 9]}
+    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0]}
     df_raw = constructor(data)
     df = nw.from_native(df_raw).lazy()
     with pytest.raises(
diff --git a/tests/expr_and_series/max_test.py b/tests/expr_and_series/max_test.py
index 09483cb7d4..f2a3196102 100644
--- a/tests/expr_and_series/max_test.py
+++ b/tests/expr_and_series/max_test.py
@@ -7,7 +7,7 @@
 from tests.utils import ConstructorEager
 from tests.utils import assert_equal_data
 
-data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8, 9]}
+data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0]}
 
 
 @pytest.mark.parametrize("expr", [nw.col("a", "b", "z").max(), nw.max("a", "b", "z")])
diff --git a/tests/expr_and_series/mean_test.py b/tests/expr_and_series/mean_test.py
index bab1fe821d..6e5ce59aff 100644
--- a/tests/expr_and_series/mean_test.py
+++ b/tests/expr_and_series/mean_test.py
@@ -7,7 +7,7 @@
 from tests.utils import ConstructorEager
 from tests.utils import assert_equal_data
 
-data = {"a": [1, 3, 2], "b": [4, 4, 7], "z": [7.0, 8, 9]}
+data = {"a": [1, 3, 2], "b": [4, 4, 7], "z": [7.0, 8.0, 9.0]}
 
 
 @pytest.mark.parametrize("expr", [nw.col("a", "b", "z").mean(), nw.mean("a", "b", "z")])
diff --git a/tests/expr_and_series/median_test.py b/tests/expr_and_series/median_test.py
index 9c509a1825..a548754b60 100644
--- a/tests/expr_and_series/median_test.py
+++ b/tests/expr_and_series/median_test.py
@@ -11,7 +11,7 @@
 data = {
     "a": [3, 8, 2, None],
     "b": [5, 5, None, 7],
-    "z": [7.0, 8, 9, None],
+    "z": [7.0, 8.0, 9.0, None],
     "s": ["f", "a", "x", "x"],
 }
 
diff --git a/tests/expr_and_series/min_test.py b/tests/expr_and_series/min_test.py
index f50facb3e8..1ed0fa2af1 100644
--- a/tests/expr_and_series/min_test.py
+++ b/tests/expr_and_series/min_test.py
@@ -7,7 +7,7 @@
 from tests.utils import ConstructorEager
 from tests.utils import assert_equal_data
 
-data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8, 9]}
+data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0]}
 
 
 @pytest.mark.parametrize("expr", [nw.col("a", "b", "z").min(), nw.min("a", "b", "z")])
diff --git a/tests/expr_and_series/n_unique_test.py b/tests/expr_and_series/n_unique_test.py
index 1bcbe89fdc..c761b20fd6 100644
--- a/tests/expr_and_series/n_unique_test.py
+++ b/tests/expr_and_series/n_unique_test.py
@@ -7,7 +7,7 @@
 
 data = {
     "a": [1.0, None, None, 3.0],
-    "b": [1.0, None, 4, 5.0],
+    "b": [1.0, None, 4.0, 5.0],
 }
 
 
diff --git a/tests/expr_and_series/nth_test.py b/tests/expr_and_series/nth_test.py
index a7dc7f6481..d2fc79b084 100644
--- a/tests/expr_and_series/nth_test.py
+++ b/tests/expr_and_series/nth_test.py
@@ -8,7 +8,7 @@
 from tests.utils import Constructor
 from tests.utils import assert_equal_data
 
-data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.1, 8, 9]}
+data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.1, 8.0, 9.0]}
 
 
 @pytest.mark.parametrize(
@@ -16,7 +16,7 @@
     [
         (0, {"a": [1, 3, 2]}),
         ([0, 1], {"a": [1, 3, 2], "b": [4, 4, 6]}),
-        ([0, 2], {"a": [1, 3, 2], "z": [7.1, 8, 9]}),
+        ([0, 2], {"a": [1, 3, 2], "z": [7.1, 8.0, 9.0]}),
     ],
 )
 def test_nth(
diff --git a/tests/expr_and_series/null_count_test.py b/tests/expr_and_series/null_count_test.py
index db162363b7..5a07f29b5f 100644
--- a/tests/expr_and_series/null_count_test.py
+++ b/tests/expr_and_series/null_count_test.py
@@ -7,7 +7,7 @@
 
 data = {
     "a": [1.0, None, None, 3.0],
-    "b": [1.0, None, 4, 5.0],
+    "b": [1.0, None, 4.0, 5.0],
 }
 
 
diff --git a/tests/expr_and_series/quantile_test.py b/tests/expr_and_series/quantile_test.py
index a9207cebd4..b18d112cfb 100644
--- a/tests/expr_and_series/quantile_test.py
+++ b/tests/expr_and_series/quantile_test.py
@@ -35,7 +35,7 @@ def test_quantile_expr(
         request.applymarker(pytest.mark.xfail)
 
     q = 0.3
-    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8, 9]}
+    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0]}
     df_raw = constructor(data)
     df = nw.from_native(df_raw)
 
@@ -71,7 +71,7 @@ def test_quantile_series(
 ) -> None:
     q = 0.3
 
-    series = nw.from_native(constructor_eager({"a": [7.0, 8, 9]}), eager_only=True)[
+    series = nw.from_native(constructor_eager({"a": [7.0, 8.0, 9.0]}), eager_only=True)[
         "a"
     ].alias("a")
     result = series.quantile(quantile=q, interpolation=interpolation)
diff --git a/tests/expr_and_series/std_test.py b/tests/expr_and_series/std_test.py
index f2eabf4f22..0cc9b67224 100644
--- a/tests/expr_and_series/std_test.py
+++ b/tests/expr_and_series/std_test.py
@@ -9,8 +9,8 @@
 from tests.utils import ConstructorEager
 from tests.utils import assert_equal_data
 
-data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8, 9]}
-data_with_nulls = {"a": [1, 3, 2, None], "b": [4, 4, 6, None], "z": [7.0, 8, 9, None]}
+data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0]}
+data_with_nulls = {"a": [1, 3, 2, None], "b": [4, 4, 6, None], "z": [7.0, 8.0, 9.0, None]}
 
 expected_results = {
     "a_ddof_1": [1.0],
diff --git a/tests/expr_and_series/sum_horizontal_test.py b/tests/expr_and_series/sum_horizontal_test.py
index decb65c028..8b3dcf5696 100644
--- a/tests/expr_and_series/sum_horizontal_test.py
+++ b/tests/expr_and_series/sum_horizontal_test.py
@@ -15,7 +15,7 @@ def test_sumh(
 ) -> None:
     if "duckdb" in str(constructor):
         request.applymarker(pytest.mark.xfail)
-    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8, 9]}
+    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0]}
     df = nw.from_native(constructor(data))
     result = df.with_columns(horizontal_sum=nw.sum_horizontal(col_expr, nw.col("b")))
     expected = {
diff --git a/tests/expr_and_series/sum_test.py b/tests/expr_and_series/sum_test.py
index f988e8991b..8f3ce950c3 100644
--- a/tests/expr_and_series/sum_test.py
+++ b/tests/expr_and_series/sum_test.py
@@ -7,7 +7,7 @@
 from tests.utils import ConstructorEager
 from tests.utils import assert_equal_data
 
-data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8, 9]}
+data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0]}
 
 
 @pytest.mark.parametrize("expr", [nw.col("a", "b", "z").sum(), nw.sum("a", "b", "z")])
diff --git a/tests/expr_and_series/unary_test.py b/tests/expr_and_series/unary_test.py
index 51d8c6361c..618a477d7f 100644
--- a/tests/expr_and_series/unary_test.py
+++ b/tests/expr_and_series/unary_test.py
@@ -17,7 +17,7 @@ def test_unary(constructor: Constructor, request: pytest.FixtureRequest) -> None
         "a": [1, 3, 2],
         "b": [4, 4, 6],
         "c": [7.0, 8.0, None],
-        "z": [7.0, 8, 9],
+        "z": [7.0, 8.0, 9.0],
     }
     result = nw.from_native(constructor(data)).select(
         a_mean=nw.col("a").mean(),
@@ -49,7 +49,7 @@ def test_unary_series(constructor_eager: ConstructorEager) -> None:
         "a": [1, 3, 2],
         "b": [4, 4, 6],
         "c": [7.0, 8.0, None],
-        "z": [7.0, 8, 9],
+        "z": [7.0, 8.0, 9.0],
     }
     df = nw.from_native(constructor_eager(data), eager_only=True)
     result = {
diff --git a/tests/expr_and_series/var_test.py b/tests/expr_and_series/var_test.py
index 2053dfe697..0edd8e305f 100644
--- a/tests/expr_and_series/var_test.py
+++ b/tests/expr_and_series/var_test.py
@@ -9,8 +9,8 @@
 from tests.utils import ConstructorEager
 from tests.utils import assert_equal_data
 
-data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8, 9]}
-data_with_nulls = {"a": [1, 3, 2, None], "b": [4, 4, 6, None], "z": [7.0, 8, 9, None]}
+data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0]}
+data_with_nulls = {"a": [1, 3, 2, None], "b": [4, 4, 6, None], "z": [7.0, 8.0, 9.0, None]}
 
 expected_results = {
     "a_ddof_1": [1.0],
diff --git a/tests/frame/add_test.py b/tests/frame/add_test.py
index e045618956..ec152f859a 100644
--- a/tests/frame/add_test.py
+++ b/tests/frame/add_test.py
@@ -10,7 +10,7 @@
 def test_add(constructor: Constructor, request: pytest.FixtureRequest) -> None:
     if "duckdb" in str(constructor):
         request.applymarker(pytest.mark.xfail)
-    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8, 9]}
+    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0]}
     df = nw.from_native(constructor(data))
     result = df.with_columns(
         c=nw.col("a") + nw.col("b"),
diff --git a/tests/frame/columns_test.py b/tests/frame/columns_test.py
index 3a18fb5917..d3b3af3497 100644
--- a/tests/frame/columns_test.py
+++ b/tests/frame/columns_test.py
@@ -12,7 +12,7 @@
 
 @pytest.mark.filterwarnings("ignore:Determining|Resolving.*")
 def test_columns(constructor: Constructor) -> None:
-    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8, 9]}
+    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0]}
     df = nw.from_native(constructor(data))
     result = df.columns
     expected = ["a", "b", "z"]
diff --git a/tests/frame/concat_test.py b/tests/frame/concat_test.py
index e38f4f4ffc..bbdd306fe3 100644
--- a/tests/frame/concat_test.py
+++ b/tests/frame/concat_test.py
@@ -12,7 +12,7 @@ def test_concat_horizontal(
 ) -> None:
     if ("pyspark" in str(constructor)) or "duckdb" in str(constructor):
         request.applymarker(pytest.mark.xfail)
-    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8, 9]}
+    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0]}
     df_left = nw.from_native(constructor(data)).lazy()
 
     data_right = {"c": [6, 12, -1], "d": [0, -4, 2]}
@@ -22,7 +22,7 @@ def test_concat_horizontal(
     expected = {
         "a": [1, 3, 2],
         "b": [4, 4, 6],
-        "z": [7.0, 8, 9],
+        "z": [7.0, 8.0, 9.0],
         "c": [6, 12, -1],
         "d": [0, -4, 2],
     }
@@ -33,7 +33,7 @@ def test_concat_horizontal(
 
 
 def test_concat_vertical(constructor: Constructor) -> None:
-    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8, 9]}
+    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0]}
     df_left = (
         nw.from_native(constructor(data)).lazy().rename({"a": "c", "b": "d"}).drop("z")
     )
diff --git a/tests/frame/double_test.py b/tests/frame/double_test.py
index a99cef5e25..93884b9dfb 100644
--- a/tests/frame/double_test.py
+++ b/tests/frame/double_test.py
@@ -6,7 +6,7 @@
 
 
 def test_double(constructor: Constructor) -> None:
-    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8, 9]}
+    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0]}
     df = nw.from_native(constructor(data))
 
     result = df.with_columns(nw.all() * 2)
diff --git a/tests/frame/drop_test.py b/tests/frame/drop_test.py
index eb9bb2660d..2ef3b4f4ab 100644
--- a/tests/frame/drop_test.py
+++ b/tests/frame/drop_test.py
@@ -24,7 +24,7 @@
     ],
 )
 def test_drop(constructor: Constructor, to_drop: list[str], expected: list[str]) -> None:
-    data = {"abc": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8, 9]}
+    data = {"abc": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0]}
     df = nw.from_native(constructor(data))
     assert df.drop(to_drop).collect_schema().names() == expected
     if not isinstance(to_drop, str):
diff --git a/tests/frame/filter_test.py b/tests/frame/filter_test.py
index 922af0e61e..579d40134d 100644
--- a/tests/frame/filter_test.py
+++ b/tests/frame/filter_test.py
@@ -12,7 +12,7 @@
 
 
 def test_filter(constructor: Constructor) -> None:
-    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8, 9]}
+    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0]}
     df = nw.from_native(constructor(data))
     result = df.filter(nw.col("a") > 1)
     expected = {"a": [3, 2], "b": [4, 6], "z": [8.0, 9.0]}
@@ -24,7 +24,7 @@ def test_filter_with_boolean_list(
 ) -> None:
     if "duckdb" in str(constructor):
         request.applymarker(pytest.mark.xfail)
-    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8, 9]}
+    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0]}
     df = nw.from_native(constructor(data))
     context = (
         pytest.raises(TypeError, match="not supported")
@@ -38,7 +38,7 @@ def test_filter_with_boolean_list(
 
 
 def test_filter_raise_on_agg_predicate(constructor: Constructor) -> None:
-    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8, 9]}
+    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0]}
     df = nw.from_native(constructor(data))
 
     context = (
@@ -56,7 +56,7 @@ def test_filter_raise_on_agg_predicate(constructor: Constructor) -> None:
 
 
 def test_filter_raise_on_shape_mismatch(constructor: Constructor) -> None:
-    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8, 9]}
+    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0]}
     df = nw.from_native(constructor(data))
 
     context = (
diff --git a/tests/frame/head_test.py b/tests/frame/head_test.py
index e817aa4169..09b36b0741 100644
--- a/tests/frame/head_test.py
+++ b/tests/frame/head_test.py
@@ -6,7 +6,7 @@
 
 
 def test_head(constructor: Constructor) -> None:
-    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8, 9]}
+    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0]}
     expected = {"a": [1, 3], "b": [4, 4], "z": [7.0, 8.0]}
 
     df_raw = constructor(data)
diff --git a/tests/frame/invalid_test.py b/tests/frame/invalid_test.py
index 5d2f7b45a3..967481f399 100644
--- a/tests/frame/invalid_test.py
+++ b/tests/frame/invalid_test.py
@@ -10,7 +10,7 @@
 
 
 def test_invalid() -> None:
-    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8, 9]}
+    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0]}
     df = nw.from_native(pa.table({"a": [1, 2], "b": [3, 4]}))
     with pytest.raises(ValueError, match="Multi-output"):
         df.select(nw.all() + nw.all())
diff --git a/tests/frame/is_duplicated_test.py b/tests/frame/is_duplicated_test.py
index bcc8037123..481fb77239 100644
--- a/tests/frame/is_duplicated_test.py
+++ b/tests/frame/is_duplicated_test.py
@@ -6,7 +6,7 @@
 
 
 def test_is_duplicated(constructor_eager: ConstructorEager) -> None:
-    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8, 9]}
+    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0]}
     df_raw = constructor_eager(data)
     df = nw.from_native(df_raw, eager_only=True)
     result = nw.concat([df, df.head(1)]).is_duplicated()
diff --git a/tests/frame/is_empty_test.py b/tests/frame/is_empty_test.py
index 7ea6b22ad9..47c347af73 100644
--- a/tests/frame/is_empty_test.py
+++ b/tests/frame/is_empty_test.py
@@ -15,7 +15,7 @@
 def test_is_empty(
     constructor_eager: ConstructorEager, threshold: Any, expected: Any
 ) -> None:
-    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8, 9]}
+    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0]}
     df_raw = constructor_eager(data)
     df = nw.from_native(df_raw, eager_only=True)
     result = df.filter(nw.col("a") > threshold).is_empty()
diff --git a/tests/frame/is_unique_test.py b/tests/frame/is_unique_test.py
index 81718f36cf..88dacd01cf 100644
--- a/tests/frame/is_unique_test.py
+++ b/tests/frame/is_unique_test.py
@@ -6,7 +6,7 @@
 
 
 def test_is_unique(constructor_eager: ConstructorEager) -> None:
-    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8, 9]}
+    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0]}
     df_raw = constructor_eager(data)
     df = nw.from_native(df_raw, eager_only=True)
     result = nw.concat([df, df.head(1)]).is_unique()
diff --git a/tests/frame/item_test.py b/tests/frame/item_test.py
index 5a5f037f1c..3646c20341 100644
--- a/tests/frame/item_test.py
+++ b/tests/frame/item_test.py
@@ -20,7 +20,7 @@ def test_item(
     column: int | str | None,
     expected: Any,
 ) -> None:
-    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8, 9]}
+    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0]}
     df = nw.from_native(constructor_eager(data), eager_only=True)
     assert_equal_data({"a": [df.item(row, column)]}, {"a": [expected]})
     assert_equal_data({"a": [df.select("a").head(1).item()]}, {"a": [1]})
@@ -52,6 +52,6 @@ def test_item_value_error(
     column: int | str | None,
     err_msg: str,
 ) -> None:
-    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8, 9]}
+    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0]}
     with pytest.raises(ValueError, match=err_msg):
         nw.from_native(constructor_eager(data), eager_only=True).item(row, column)
diff --git a/tests/frame/join_test.py b/tests/frame/join_test.py
index 4c3c1419b5..68e2ed8b8d 100644
--- a/tests/frame/join_test.py
+++ b/tests/frame/join_test.py
@@ -21,7 +21,7 @@ def test_inner_join_two_keys(constructor: Constructor) -> None:
     data = {
         "antananarivo": [1, 3, 2],
         "bob": [4, 4, 6],
-        "zor ro": [7.0, 8, 9],
+        "zor ro": [7.0, 8.0, 9.0],
         "idx": [0, 1, 2],
     }
     df = nw_main.from_native(constructor(data))
@@ -38,9 +38,9 @@ def test_inner_join_two_keys(constructor: Constructor) -> None:
     expected = {
         "antananarivo": [1, 3, 2],
         "bob": [4, 4, 6],
-        "zor ro": [7.0, 8, 9],
+        "zor ro": [7.0, 8.0, 9.0],
         "idx": [0, 1, 2],
-        "zor ro_right": [7.0, 8, 9],
+        "zor ro_right": [7.0, 8.0, 9.0],
     }
     assert_equal_data(result, expected)
     assert_equal_data(result_on, expected)
@@ -50,7 +50,7 @@ def test_inner_join_single_key(constructor: Constructor) -> None:
     data = {
         "antananarivo": [1, 3, 2],
         "bob": [4, 4, 6],
-        "zor ro": [7.0, 8, 9],
+        "zor ro": [7.0, 8.0, 9.0],
         "idx": [0, 1, 2],
     }
     df = nw.from_native(constructor(data))
@@ -67,10 +67,10 @@ def test_inner_join_single_key(constructor: Constructor) -> None:
     expected = {
         "antananarivo": [1, 3, 2],
         "bob": [4, 4, 6],
-        "zor ro": [7.0, 8, 9],
+        "zor ro": [7.0, 8.0, 9.0],
         "idx": [0, 1, 2],
         "bob_right": [4, 4, 6],
-        "zor ro_right": [7.0, 8, 9],
+        "zor ro_right": [7.0, 8.0, 9.0],
     }
     assert_equal_data(result, expected)
     assert_equal_data(result_on, expected)
@@ -100,7 +100,7 @@ def test_suffix(constructor: Constructor, how: str, suffix: str) -> None:
     data = {
         "antananarivo": [1, 3, 2],
         "bob": [4, 4, 6],
-        "zor ro": [7.0, 8, 9],
+        "zor ro": [7.0, 8.0, 9.0],
     }
     df = nw.from_native(constructor(data))
     df_right = df
@@ -168,7 +168,7 @@ def test_anti_join(
     filter_expr: nw.Expr,
     expected: dict[str, list[Any]],
 ) -> None:
-    data = {"antananarivo": [1, 3, 2], "bob": [4, 4, 6], "zor ro": [7.0, 8, 9]}
+    data = {"antananarivo": [1, 3, 2], "bob": [4, 4, 6], "zor ro": [7.0, 8.0, 9.0]}
     df = nw.from_native(constructor(data))
     other = df.filter(filter_expr)
     result = df.join(other, how="anti", left_on=join_key, right_on=join_key)  # type: ignore[arg-type]
@@ -206,7 +206,7 @@ def test_semi_join(
     filter_expr: nw.Expr,
     expected: dict[str, list[Any]],
 ) -> None:
-    data = {"antananarivo": [1, 3, 2], "bob": [4, 4, 6], "zor ro": [7.0, 8, 9]}
+    data = {"antananarivo": [1, 3, 2], "bob": [4, 4, 6], "zor ro": [7.0, 8.0, 9.0]}
     df = nw.from_native(constructor(data))
     other = df.filter(filter_expr)
     result = df.join(other, how="semi", left_on=join_key, right_on=join_key).sort(  # type: ignore[arg-type]
@@ -217,7 +217,7 @@ def test_semi_join(
 
 @pytest.mark.parametrize("how", ["right", "full"])
 def test_join_not_implemented(constructor: Constructor, how: str) -> None:
-    data = {"antananarivo": [1, 3, 2], "bob": [4, 4, 6], "zor ro": [7.0, 8, 9]}
+    data = {"antananarivo": [1, 3, 2], "bob": [4, 4, 6], "zor ro": [7.0, 8.0, 9.0]}
     df = nw.from_native(constructor(data))
 
     with pytest.raises(
@@ -231,13 +231,13 @@ def test_join_not_implemented(constructor: Constructor, how: str) -> None:
 
 def test_left_join(constructor: Constructor) -> None:
     data_left = {
-        "antananarivo": [1.0, 2, 3],
-        "bob": [4.0, 5, 6],
+        "antananarivo": [1.0, 2.0, 3.0],
+        "bob": [4.0, 5.0, 6.0],
         "idx": [0.0, 1.0, 2.0],
     }
     data_right = {
-        "antananarivo": [1.0, 2, 3],
-        "co": [4.0, 5, 7],
+        "antananarivo": [1.0, 2.0, 3.0],
+        "co": [4.0, 5.0, 7.0],
         "idx": [0.0, 1.0, 2.0],
     }
     df_left = nw.from_native(constructor(data_left))
@@ -286,15 +286,15 @@ def test_left_join_multiple_column(constructor: Constructor) -> None:
 
 def test_left_join_overlapping_column(constructor: Constructor) -> None:
     data_left = {
-        "antananarivo": [1.0, 2, 3],
-        "bob": [4.0, 5, 6],
-        "d": [1.0, 4, 2],
+        "antananarivo": [1.0, 2.0, 3.0],
+        "bob": [4.0, 5.0, 6.0],
+        "d": [1.0, 4.0, 2.0],
         "idx": [0.0, 1.0, 2.0],
     }
     data_right = {
-        "antananarivo": [1.0, 2, 3],
-        "c": [4.0, 5, 6],
-        "d": [1.0, 4, 2],
+        "antananarivo": [1.0, 2.0, 3.0],
+        "c": [4.0, 5.0, 6.0],
+        "d": [1.0, 4.0, 2.0],
         "idx": [0.0, 1.0, 2.0],
     }
     df_left = nw.from_native(constructor(data_left))
@@ -331,7 +331,7 @@ def test_left_join_overlapping_column(constructor: Constructor) -> None:
 
 @pytest.mark.parametrize("how", ["inner", "left", "semi", "anti"])
 def test_join_keys_exceptions(constructor: Constructor, how: str) -> None:
-    data = {"antananarivo": [1, 3, 2], "bob": [4, 4, 6], "zor ro": [7.0, 8, 9]}
+    data = {"antananarivo": [1, 3, 2], "bob": [4, 4, 6], "zor ro": [7.0, 8.0, 9.0]}
     df = nw.from_native(constructor(data))
 
     with pytest.raises(
@@ -536,7 +536,7 @@ def test_joinasof_by(
 def test_joinasof_not_implemented(
     constructor: Constructor, strategy: Literal["backward", "forward"]
 ) -> None:
-    data = {"antananarivo": [1, 3, 2], "bob": [4, 4, 6], "zor ro": [7.0, 8, 9]}
+    data = {"antananarivo": [1, 3, 2], "bob": [4, 4, 6], "zor ro": [7.0, 8.0, 9.0]}
     df = nw.from_native(constructor(data))
 
     with pytest.raises(
@@ -552,7 +552,7 @@ def test_joinasof_not_implemented(
 
 
 def test_joinasof_keys_exceptions(constructor: Constructor) -> None:
-    data = {"antananarivo": [1, 3, 2], "bob": [4, 4, 6], "zor ro": [7.0, 8, 9]}
+    data = {"antananarivo": [1, 3, 2], "bob": [4, 4, 6], "zor ro": [7.0, 8.0, 9.0]}
     df = nw.from_native(constructor(data))
 
     with pytest.raises(
@@ -593,7 +593,7 @@ def test_joinasof_keys_exceptions(constructor: Constructor) -> None:
 
 
 def test_joinasof_by_exceptions(constructor: Constructor) -> None:
-    data = {"antananarivo": [1, 3, 2], "bob": [4, 4, 6], "zor ro": [7.0, 8, 9]}
+    data = {"antananarivo": [1, 3, 2], "bob": [4, 4, 6], "zor ro": [7.0, 8.0, 9.0]}
     df = nw.from_native(constructor(data))
     with pytest.raises(
         ValueError,
diff --git a/tests/frame/reindex_test.py b/tests/frame/reindex_test.py
index e31ca93632..6cad175c4f 100644
--- a/tests/frame/reindex_test.py
+++ b/tests/frame/reindex_test.py
@@ -8,7 +8,7 @@
 import narwhals.stable.v1 as nw
 from tests.utils import assert_equal_data
 
-data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8, 9]}
+data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0]}
 
 
 @pytest.mark.parametrize("df_raw", [pd.DataFrame(data)])
diff --git a/tests/frame/rename_test.py b/tests/frame/rename_test.py
index 24c0462002..ccc832ebb8 100644
--- a/tests/frame/rename_test.py
+++ b/tests/frame/rename_test.py
@@ -6,8 +6,8 @@
 
 
 def test_rename(constructor: Constructor) -> None:
-    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8, 9]}
+    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0]}
     df = nw.from_native(constructor(data))
     result = df.rename({"a": "x", "b": "y"})
-    expected = {"x": [1, 3, 2], "y": [4, 4, 6], "z": [7.0, 8, 9]}
+    expected = {"x": [1, 3, 2], "y": [4, 4, 6], "z": [7.0, 8.0, 9.0]}
     assert_equal_data(result, expected)
diff --git a/tests/frame/rows_test.py b/tests/frame/rows_test.py
index cdf426483a..a2a70f46bf 100644
--- a/tests/frame/rows_test.py
+++ b/tests/frame/rows_test.py
@@ -11,7 +11,7 @@
 if TYPE_CHECKING:
     from tests.utils import ConstructorEager
 
-data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8, 9]}
+data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0]}
 data_na = {"a": [None, 3, 2], "b": [4, 4, 6], "z": [7.0, None, 9]}
 
 
@@ -38,7 +38,7 @@ def test_iter_rows(
     if "cudf" in str(constructor_eager):
         request.applymarker(pytest.mark.xfail)
 
-    data = {"a": [1, 3, 2], "_b": [4, 4, 6], "z": [7.0, 8, 9], "1": [5, 6, 7]}
+    data = {"a": [1, 3, 2], "_b": [4, 4, 6], "z": [7.0, 8.0, 9.0], "1": [5, 6, 7]}
     df = nw.from_native(constructor_eager(data), eager_only=True)
     result = list(df.iter_rows(named=named))
     assert result == expected
@@ -94,7 +94,7 @@ def test_rows_eager(
     expected: list[tuple[Any, ...]] | list[dict[str, Any]],
 ) -> None:
     # posit-dev/py-shiny relies on `.rows(named=False)` to return unnamed rows
-    data = {"a": [1, 3, 2], "_b": [4, 4, 6], "z": [7.0, 8, 9], "1": [5, 6, 7]}
+    data = {"a": [1, 3, 2], "_b": [4, 4, 6], "z": [7.0, 8.0, 9.0], "1": [5, 6, 7]}
     df = nw.from_native(constructor_eager(data), eager_only=True)
     result = df.rows(named=named)
     assert result == expected
diff --git a/tests/frame/schema_test.py b/tests/frame/schema_test.py
index 565bf01597..9b395aaa80 100644
--- a/tests/frame/schema_test.py
+++ b/tests/frame/schema_test.py
@@ -27,7 +27,9 @@
 
 @pytest.mark.filterwarnings("ignore:Determining|Resolving.*")
 def test_schema(constructor: Constructor) -> None:
-    df = nw.from_native(constructor({"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.1, 8, 9]}))
+    df = nw.from_native(
+        constructor({"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.1, 8.0, 9.0]})
+    )
     result = df.schema
     expected = {"a": nw.Int64, "b": nw.Int64, "z": nw.Float64}
 
@@ -38,7 +40,9 @@ def test_schema(constructor: Constructor) -> None:
 
 
 def test_collect_schema(constructor: Constructor) -> None:
-    df = nw.from_native(constructor({"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.1, 8, 9]}))
+    df = nw.from_native(
+        constructor({"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.1, 8.0, 9.0]})
+    )
     expected = {"a": nw.Int64, "b": nw.Int64, "z": nw.Float64}
 
     result = df.collect_schema()
diff --git a/tests/frame/select_test.py b/tests/frame/select_test.py
index 946e582030..3593eed2ce 100644
--- a/tests/frame/select_test.py
+++ b/tests/frame/select_test.py
@@ -20,7 +20,7 @@ class Foo: ...
 
 
 def test_select(constructor: Constructor) -> None:
-    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8, 9]}
+    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0]}
     df = nw.from_native(constructor(data))
     result = df.select("a")
     expected = {"a": [1, 3, 2]}
@@ -82,7 +82,7 @@ def test_missing_columns(
 ) -> None:
     if ("pyspark" in str(constructor)) or "duckdb" in str(constructor):
         request.applymarker(pytest.mark.xfail)
-    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8, 9]}
+    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0]}
     df = nw.from_native(constructor(data))
     selected_columns = ["a", "e", "f"]
     msg = (
diff --git a/tests/frame/sort_test.py b/tests/frame/sort_test.py
index 1ce3414c8e..5833b5ccc3 100644
--- a/tests/frame/sort_test.py
+++ b/tests/frame/sort_test.py
@@ -8,7 +8,7 @@
 
 
 def test_sort(constructor: Constructor) -> None:
-    data = {"an tan": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8, 9]}
+    data = {"an tan": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0]}
     df = nw.from_native(constructor(data))
     result = df.sort("an tan", "b")
     expected = {
diff --git a/tests/frame/tail_test.py b/tests/frame/tail_test.py
index 75f46a4a13..c261477f91 100644
--- a/tests/frame/tail_test.py
+++ b/tests/frame/tail_test.py
@@ -13,7 +13,7 @@ def test_tail(request: pytest.FixtureRequest, constructor: Constructor) -> None:
     if "pyspark" in str(constructor):
         request.applymarker(pytest.mark.xfail)
 
-    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8, 9]}
+    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0]}
     expected = {"a": [3, 2], "b": [4, 6], "z": [8.0, 9]}
 
     df_raw = constructor(data)
diff --git a/tests/frame/to_arrow_test.py b/tests/frame/to_arrow_test.py
index 70913ed155..651dfabad1 100644
--- a/tests/frame/to_arrow_test.py
+++ b/tests/frame/to_arrow_test.py
@@ -21,7 +21,7 @@ def test_to_arrow(
         # pyarrow requires pandas>=1.0.0
         request.applymarker(pytest.mark.xfail)
 
-    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.1, 8, 9]}
+    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.1, 8.0, 9.0]}
     df_raw = constructor_eager(data)
     result = nw.from_native(df_raw, eager_only=True).to_arrow()
 
diff --git a/tests/frame/to_dict_test.py b/tests/frame/to_dict_test.py
index e6a434b7f2..fa31a7a725 100644
--- a/tests/frame/to_dict_test.py
+++ b/tests/frame/to_dict_test.py
@@ -11,14 +11,14 @@
     "ignore:.*all arguments of to_dict except for the argument:FutureWarning"
 )
 def test_to_dict(constructor_eager: ConstructorEager) -> None:
-    data = {"a": [1, 3, 2], "b": [4, 4, 6], "c": [7.0, 8, 9]}
+    data = {"a": [1, 3, 2], "b": [4, 4, 6], "c": [7.0, 8.0, 9.0]}
     df = nw.from_native(constructor_eager(data), eager_only=True)
     result = df.to_dict(as_series=False)
     assert result == data
 
 
 def test_to_dict_as_series(constructor_eager: ConstructorEager) -> None:
-    data = {"a": [1, 3, 2], "b": [4, 4, 6], "c": [7.0, 8, 9]}
+    data = {"a": [1, 3, 2], "b": [4, 4, 6], "c": [7.0, 8.0, 9.0]}
     df = nw.from_native(constructor_eager(data), eager_only=True)
     result = df.to_dict(as_series=True)
     assert isinstance(result["a"], nw.Series)
diff --git a/tests/frame/to_native_test.py b/tests/frame/to_native_test.py
index fb90caf107..08cf8efd76 100644
--- a/tests/frame/to_native_test.py
+++ b/tests/frame/to_native_test.py
@@ -9,7 +9,7 @@
 
 
 def test_to_native(constructor: Constructor) -> None:
-    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.1, 8, 9]}
+    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.1, 8.0, 9.0]}
     df_raw = constructor(data)
     df = nw.from_native(df_raw)
 
diff --git a/tests/frame/to_numpy_test.py b/tests/frame/to_numpy_test.py
index e1179f3a57..ca9d837e96 100644
--- a/tests/frame/to_numpy_test.py
+++ b/tests/frame/to_numpy_test.py
@@ -16,11 +16,11 @@
 
 
 def test_to_numpy(constructor_eager: ConstructorEager) -> None:
-    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.1, 8, 9]}
+    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.1, 8.0, 9.0]}
     df_raw = constructor_eager(data)
     result = nw.from_native(df_raw, eager_only=True).to_numpy()
 
-    expected = np.array([[1, 3, 2], [4, 4, 6], [7.1, 8, 9]]).T
+    expected = np.array([[1, 3, 2], [4, 4, 6], [7.1, 8.0, 9.0]]).T
     np.testing.assert_array_equal(result, expected)
     assert result.dtype == "float64"
 
diff --git a/tests/frame/to_pandas_test.py b/tests/frame/to_pandas_test.py
index 1bc588f353..07b35c5872 100644
--- a/tests/frame/to_pandas_test.py
+++ b/tests/frame/to_pandas_test.py
@@ -20,7 +20,7 @@
 def test_convert_pandas(
     constructor_eager: ConstructorEager,
 ) -> None:
-    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8, 9]}
+    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0]}
     df_raw = constructor_eager(data)
     result = nw.from_native(df_raw).to_pandas()  # type: ignore[union-attr]
 
diff --git a/tests/frame/to_polars_test.py b/tests/frame/to_polars_test.py
index d8683cbd0e..cc2f860d99 100644
--- a/tests/frame/to_polars_test.py
+++ b/tests/frame/to_polars_test.py
@@ -14,7 +14,7 @@
 
 @pytest.mark.filterwarnings("ignore:.*Passing a BlockManager.*:DeprecationWarning")
 def test_convert_polars(constructor_eager: ConstructorEager) -> None:
-    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.1, 8, 9]}
+    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.1, 8.0, 9.0]}
     df_raw = constructor_eager(data)
     result = nw.from_native(df_raw).to_polars()  # type: ignore[union-attr]
 
diff --git a/tests/frame/unique_test.py b/tests/frame/unique_test.py
index a193ab98b5..8e1b71de0e 100644
--- a/tests/frame/unique_test.py
+++ b/tests/frame/unique_test.py
@@ -12,7 +12,7 @@
 from tests.utils import Constructor
 from tests.utils import assert_equal_data
 
-data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8, 9]}
+data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0]}
 
 
 @pytest.mark.parametrize("subset", ["b", ["b"]])
diff --git a/tests/frame/with_columns_test.py b/tests/frame/with_columns_test.py
index 6fa3ab8258..7badea6a11 100644
--- a/tests/frame/with_columns_test.py
+++ b/tests/frame/with_columns_test.py
@@ -22,23 +22,23 @@ def test_with_columns_int_col_name_pandas() -> None:
 
 
 def test_with_columns_order(constructor: Constructor) -> None:
-    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8, 9]}
+    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0]}
     df = nw.from_native(constructor(data))
     result = df.with_columns(nw.col("a") + 1, d=nw.col("a") - 1)
     assert result.collect_schema().names() == ["a", "b", "z", "d"]
-    expected = {"a": [2, 4, 3], "b": [4, 4, 6], "z": [7.0, 8, 9], "d": [0, 2, 1]}
+    expected = {"a": [2, 4, 3], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0], "d": [0, 2, 1]}
     assert_equal_data(result, expected)
 
 
 def test_with_columns_empty(constructor: Constructor) -> None:
-    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8, 9]}
+    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0]}
     df = nw.from_native(constructor(data))
     result = df.select().with_columns()
     assert_equal_data(result, {})
 
 
 def test_with_columns_order_single_row(constructor: Constructor) -> None:
-    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8, 9], "i": [0, 1, 2]}
+    data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0], "i": [0, 1, 2]}
     df = nw.from_native(constructor(data)).filter(nw.col("i") < 1).drop("i")
     result = df.with_columns(nw.col("a") + 1, d=nw.col("a") - 1)
     assert result.collect_schema().names() == ["a", "b", "z", "d"]
diff --git a/tests/group_by_test.py b/tests/group_by_test.py
index 07ab957451..586b75dc89 100644
--- a/tests/group_by_test.py
+++ b/tests/group_by_test.py
@@ -15,7 +15,7 @@
 from tests.utils import ConstructorEager
 from tests.utils import assert_equal_data
 
-data = {"a": [1, 1, 3], "b": [4, 4, 6], "c": [7.0, 8, 9]}
+data = {"a": [1, 1, 3], "b": [4, 4, 6], "c": [7.0, 8.0, 9.0]}
 
 df_pandas = pd.DataFrame(data)
 df_lazy = pl.LazyFrame(data)
diff --git a/tests/series_only/is_sorted_test.py b/tests/series_only/is_sorted_test.py
index 2ff6e50f1f..7fa0294952 100644
--- a/tests/series_only/is_sorted_test.py
+++ b/tests/series_only/is_sorted_test.py
@@ -8,7 +8,7 @@
 
 data = [1, 3, 2]
 data_dups = [4, 4, 6]
-data_sorted = [7.0, 8, 9]
+data_sorted = [7.0, 8.0, 9.0]
 
 
 @pytest.mark.parametrize(