ci: use duckdb instead of ibis to test interchange-only support

altair/utils/data.py

@@ -65,8 +65,9 @@ class SupportsGeoInterface(Protocol):
 
 
 def is_data_type(obj: Any) -> TypeIs[DataType]:
-    return _is_pandas_dataframe(obj) or isinstance(
-        obj, (dict, DataFrameLike, SupportsGeoInterface, nw.DataFrame)
+    return isinstance(obj, (dict, SupportsGeoInterface)) or isinstance(
+        nw.from_native(obj, eager_or_interchange_only=True, strict=False),
+        nw.DataFrame,
     )
 
 

pyproject.toml

@@ -69,7 +69,7 @@ all = [
 dev = [
     "hatch",
     "ruff>=0.6.0",
-    "ibis-framework[polars]",
+    "duckdb>=1.0",
     "ipython[kernel]",
     "pandas>=1.1.3",
     "pytest",

tests/vegalite/v5/test_api.py

@@ -15,7 +15,7 @@
 from importlib.util import find_spec
 from typing import TYPE_CHECKING
 
-import ibis
+import duckdb
 import jsonschema
 import narwhals.stable.v1 as nw
 import pandas as pd
@@ -26,7 +26,7 @@
 import altair as alt
 from altair.utils.core import use_signature
 from altair.utils.schemapi import Optional, SchemaValidationError, Undefined
-from tests import skip_requires_vl_convert, slow
+from tests import skip_requires_pyarrow, skip_requires_vl_convert, slow
 
 if TYPE_CHECKING:
     from typing import Any
@@ -1607,51 +1607,72 @@ def test_polars_with_pandas_nor_pyarrow(monkeypatch: pytest.MonkeyPatch):
     assert "numpy" not in sys.modules
 
 
-@pytest.mark.skipif(
-    Version("1.5") > PANDAS_VERSION,
-    reason="A warning is thrown on old pandas versions",
-)
-@pytest.mark.xfail(
-    sys.platform == "win32", reason="Timezone database is not installed on Windows"
-)
-def test_ibis_with_date_32():
-    ibis.set_backend("polars")
-    df = pl.DataFrame(
+@skip_requires_pyarrow
+def test_interchange_with_date_32():
+    # Test that objects which Narwhals only supports at the interchange
+    # level can be plotted when they contain date32 columns.
+    df = pl.DataFrame(  # noqa: F841
         {"a": [1, 2, 3], "b": [date(2020, 1, 1), date(2020, 1, 2), date(2020, 1, 3)]}
     )
-    tbl = ibis.memtable(df)
-    result = alt.Chart(tbl).mark_line().encode(x="a", y="b").to_dict()
+    rel = duckdb.sql("select * from df")
+    result = alt.Chart(rel).mark_line().encode(x="a", y="b").to_dict()
     assert next(iter(result["datasets"].values())) == [
         {"a": 1, "b": "2020-01-01T00:00:00"},
         {"a": 2, "b": "2020-01-02T00:00:00"},
         {"a": 3, "b": "2020-01-03T00:00:00"},
     ]
 
 
-@pytest.mark.skipif(
-    Version("1.5") > PANDAS_VERSION,
-    reason="A warning is thrown on old pandas versions",
-)
-@pytest.mark.xfail(
-    sys.platform == "win32", reason="Timezone database is not installed on Windows"
-)
-def test_ibis_with_vegafusion(monkeypatch: pytest.MonkeyPatch):
-    ibis.set_backend("polars")
-    df = pl.DataFrame(
+@skip_requires_pyarrow
+def test_interchange_with_vegafusion(monkeypatch: pytest.MonkeyPatch):
+    # Test that objects which Narwhals only supports at the interchange
+    # level don't get converted to PyArrow unnecessarily when plotted
+    # with the vegafusion transformer.
+    # TODO: this test can be drastically simplified when some level of
+    # DuckDB support in VegaFusion, as it can then just be `alt.Chart(rel_df)`
+    # without DuckDBWithInterchangeSupport.
+    df = pl.DataFrame(  # noqa: F841
         {
             "a": [1, 2, 3],
             "b": [datetime(2020, 1, 1), datetime(2020, 1, 2), datetime(2020, 1, 3)],
         }
     )
-    tbl = ibis.memtable(df)
+    rel = duckdb.sql("select * from df")
+
+    class DuckDBWithInterchangeSupport:
+        """
+        DuckDB doesn't (yet?) support the interchange protocol.
+
+        So, we create duckdb wrapper which defers to PyArrow's
+        implementation of the protocol.
+        """
+
+        def __init__(self, rel: duckdb.DuckDBPyRelation) -> None:
+            self._rel = rel
+
+        def __dataframe__(self, allow_copy: bool = True) -> object:
+            return self._rel.to_arrow_table().__dataframe__()
+
+    rel_df = DuckDBWithInterchangeSupport(rel)
     # "poison" `arrow_table_from_dfi_dataframe` to check that it does not get called
     # if we use the vegafusion transformer
     monkeypatch.setattr(
         "altair.utils.data.arrow_table_from_dfi_dataframe", lambda x: 1 / 0
     )
-    tbl = ibis.memtable(df)
+
+    # Narwhals doesn't fully support our custom DuckDBWithInterchangeSupport,
+    # so we need to overwrite `to_native`
+    def to_native(df, strict):
+        if isinstance(df, nw.DataFrame):
+            return rel_df
+        return df
+
+    monkeypatch.setattr("narwhals.stable.v1.to_native", to_native)
+
     with alt.data_transformers.enable("vegafusion"):
-        result = alt.Chart(tbl).mark_line().encode(x="a", y="b").to_dict(format="vega")
+        result = (
+            alt.Chart(rel_df).mark_line().encode(x="a", y="b").to_dict(format="vega")
+        )
     assert next(iter(result["data"]))["values"] == [
         {"a": 1, "b": "2020-01-01T00:00:00.000"},
         {"a": 2, "b": "2020-01-02T00:00:00.000"},