Test additional long-form data parsing functionality

seaborn/tests/_core/test_data.py

@@ -19,7 +19,7 @@ def long_variables(self):
         variables = dict(x="x", y="y", hue="a", size="z", style="s_cat")
         return variables
 
-    def test_long_df(self, long_df, long_variables):
+    def test_named_vectors(self, long_df, long_variables):
 
         p = PlotData(long_df, long_variables)
         assert p._source_data is long_df
@@ -28,7 +28,7 @@ def test_long_df(self, long_df, long_variables):
             assert p.names[key] == val
             assert_series_equal(p.frame[key], long_df[val])
 
-    def test_long_df_and_vectors(self, long_df, long_variables):
+    def test_named_and_given_vectors(self, long_df, long_variables):
 
         long_variables["y"] = long_df["b"]
         long_variables["size"] = long_df["z"].to_numpy()
@@ -43,7 +43,7 @@ def test_long_df_and_vectors(self, long_df, long_variables):
         assert p.names["y"] == "b"
         assert p.names["size"] is None
 
-    def test_long_df_with_index(self, long_df, long_variables):
+    def test_index_as_variable(self, long_df, long_variables):
 
         index = pd.Int64Index(np.arange(len(long_df)) * 2 + 10, name="i")
         long_variables["x"] = "i"
@@ -52,7 +52,7 @@ def test_long_df_with_index(self, long_df, long_variables):
         assert p.names["x"] == "i"
         assert_series_equal(p.frame["x"], pd.Series(index, index))
 
-    def test_long_df_with_multiindex(self, long_df, long_variables):
+    def test_multiindex_as_variables(self, long_df, long_variables):
 
         index_i = pd.Int64Index(np.arange(len(long_df)) * 2 + 10, name="i")
         index_j = pd.Int64Index(np.arange(len(long_df)) * 3 + 5, name="j")
@@ -63,7 +63,35 @@ def test_long_df_with_multiindex(self, long_df, long_variables):
         assert_series_equal(p.frame["x"], pd.Series(index_i, index))
         assert_series_equal(p.frame["y"], pd.Series(index_j, index))
 
-    def test_long_dict(self, long_dict, long_variables):
+    def test_int_as_variable_key(self):
+
+        df = pd.DataFrame(np.random.uniform(size=(10, 3)))
+
+        var = "x"
+        key = 2
+
+        p = PlotData(df, {var: key})
+        assert_series_equal(p.frame[var], df[key])
+        assert p.names[var] == str(key)
+
+    def test_int_as_variable_value(self, long_df):
+
+        p = PlotData(long_df, {"x": 0, "y": "y"})
+        assert (p.frame["x"] == 0).all()
+        assert p.names["x"] is None
+
+    def test_tuple_as_variable_key(self):
+
+        cols = pd.MultiIndex.from_product([("a", "b", "c"), ("x", "y")])
+        df = pd.DataFrame(np.random.uniform(size=(10, 6)), columns=cols)
+
+        var = "hue"
+        key = ("b", "y")
+        p = PlotData(df, {var: key})
+        assert_series_equal(p.frame[var], df[key])
+        assert p.names[var] == str(key)
+
+    def test_dict_as_data(self, long_dict, long_variables):
 
         p = PlotData(long_dict, long_variables)
         assert p._source_data is long_dict
@@ -74,7 +102,7 @@ def test_long_dict(self, long_dict, long_variables):
         "vector_type",
         ["series", "numpy", "list"],
     )
-    def test_long_vectors(self, long_df, long_variables, vector_type):
+    def test_vectors_various_types(self, long_df, long_variables, vector_type):
 
         variables = {key: long_df[val] for key, val in long_variables.items()}
         if vector_type == "numpy":
@@ -103,17 +131,6 @@ def test_none_as_variable_value(self, long_df):
         assert list(p.frame.columns) == ["x"]
         assert p.names == {"x": "z"}
 
-    def test_long_undefined_variables(self, long_df):
-
-        with pytest.raises(ValueError):
-            PlotData(long_df, dict(x="not_in_df"))
-
-        with pytest.raises(ValueError):
-            PlotData(long_df, dict(x="x", y="not_in_df"))
-
-        with pytest.raises(ValueError):
-            PlotData(long_df, dict(x="x", y="y", hue="not_in_df"))
-
     def test_frame_and_vector_mismatched_lengths(self, long_df):
 
         vector = np.arange(len(long_df) * 2)
@@ -169,7 +186,41 @@ def test_index_alignment_between_series(self):
         assert_series_equal(p.frame["x"], x_col_expected)
         assert_series_equal(p.frame["y"], y_col_expected)
 
-    def test_contains(self, long_df):
+    def test_key_not_in_data_raises(self, long_df):
+
+        var = "x"
+        key = "what"
+        msg = f"Could not interpret value `{key}` for parameter `{var}`"
+        with pytest.raises(ValueError, match=msg):
+            PlotData(long_df, {var: key})
+
+    def test_key_with_no_data_raises(self):
+
+        var = "x"
+        key = "what"
+        msg = f"Could not interpret value `{key}` for parameter `{var}`"
+        with pytest.raises(ValueError, match=msg):
+            PlotData(variables={var: key})
+
+    def test_data_vector_different_lengths_raises(self, long_df):
+
+        vector = np.arange(len(long_df) - 5)
+        msg = "Length of ndarray vectors must match length of `data`"
+        with pytest.raises(ValueError, match=msg):
+            PlotData(long_df, {"y": vector})
+
+    def test_undefined_variables_raise(self, long_df):
+
+        with pytest.raises(ValueError):
+            PlotData(long_df, dict(x="not_in_df"))
+
+        with pytest.raises(ValueError):
+            PlotData(long_df, dict(x="x", y="not_in_df"))
+
+        with pytest.raises(ValueError):
+            PlotData(long_df, dict(x="x", y="y", hue="not_in_df"))
+
+    def test_contains_operation(self, long_df):
 
         p = PlotData(long_df, {"x": "y", "hue": long_df["a"]})
         assert "x" in p