#1017 Fix is_column_of_iterables to check all values instead of just first row

cdisc_rules_engine/check_operators/dataframe_operators.py

@@ -153,8 +153,11 @@ def get_comparator_data(self, comparator, value_is_literal: bool = False):
 
     @log_operator_execution
     def is_column_of_iterables(self, column):
-        return self.value.is_series(column) and (
-            isinstance(column.iloc[0], list) or isinstance(column.iloc[0], set)
+        return self.value.is_series(column) and all(
+            val is not None
+            and not (isinstance(val, float) and pd.isna(val))
+            and isinstance(val, (list, set))
+            for val in column
         )
 
     @log_operator_execution

resources/schema/Operator.md

@@ -299,6 +299,8 @@ Complement of `suffix_equal_to`
 
 Will return True if the value in `value` is contained within the collection/iterable in the target column, or if there's an exact match for non-iterable data.
 
+The operator checks if every value in a column is a list or set. If yes, it compares row-by-row. If any value is blank or a different type (like a string or number), it compares each value against the entire column instead.
+
 Example:
 
 ```yaml
@@ -649,6 +651,8 @@ Testing whether individual values or string parts belong to specific lists or se
 
 Value in `name` compared against a list in `value`. The list can have literal values or be a reference to a `$variable`.
 
+This operator behaves similarly to `contains`. The key distinction: `contains` checks if comparator ∈ target, while `is_contained_by` checks if target ∈ comparator.
+
 > ACTARM in ('Screen Failure', 'Not Assigned', 'Not Treated', 'Unplanned Treatment')
 
 ```yaml

tests/unit/test_check_operators/test_containment_checks.py

@@ -417,3 +417,97 @@ def test_is_not_contained_by_case_insensitive(
         {"target": "target", "comparator": comparator}
     )
     assert result.equals(df.convert_to_series(expected_result))
+
+
+@pytest.mark.parametrize(
+    "data,comparator,dataset_type,expected_result",
+    [
+        (
+            {"target": [None, ["A", "B"], ["C", "D"]]},
+            "A",
+            PandasDataset,
+            [False, True, False],
+        ),
+        (
+            {"target": [None, ["A", "B"], ["C", "D"]]},
+            "C",
+            PandasDataset,
+            [False, False, True],
+        ),
+        (
+            {"target": [["A", "B"], None, ["C", "D"]]},
+            "A",
+            PandasDataset,
+            [True, False, False],
+        ),
+        (
+            {"target": [["A", "B"], ["C", "D"], None]},
+            "A",
+            PandasDataset,
+            [True, False, False],
+        ),
+    ],
+)
+def test_contains_with_none_first_row(data, comparator, dataset_type, expected_result):
+    df = dataset_type.from_dict(data)
+    dataframe_operator = DataframeType({"value": df})
+    result = dataframe_operator.contains(
+        {"target": "target", "comparator": comparator, "value_is_literal": True}
+    )
+    assert result.equals(df.convert_to_series(expected_result))
+
+
+@pytest.mark.parametrize(
+    "data,comparator,dataset_type,expected_result",
+    [
+        (
+            {"target": [None, ["A", "B"], ["C", "D"]]},
+            "A",
+            PandasDataset,
+            [True, False, True],
+        ),
+        (
+            {"target": [None, ["A", "B"], ["C", "D"]]},
+            "C",
+            PandasDataset,
+            [True, True, False],
+        ),
+    ],
+)
+def test_does_not_contain_with_none_first_row(
+    data, comparator, dataset_type, expected_result
+):
+    df = dataset_type.from_dict(data)
+    dataframe_operator = DataframeType({"value": df})
+    result = dataframe_operator.does_not_contain(
+        {"target": "target", "comparator": comparator, "value_is_literal": True}
+    )
+    assert result.equals(df.convert_to_series(expected_result))
+
+
+@pytest.mark.parametrize(
+    "data,comparator,dataset_type,expected_result",
+    [
+        (
+            {"target": ["A", "B", "C"], "comparison": [None, ["A", "B"], ["C", "D"]]},
+            "comparison",
+            PandasDataset,
+            [False, False, False],
+        ),
+        (
+            {"target": ["A", "B", "C"], "comparison": [["A", "B"], None, ["C", "D"]]},
+            "comparison",
+            PandasDataset,
+            [False, False, False],
+        ),
+    ],
+)
+def test_is_contained_by_with_none_in_comparison(
+    data, comparator, dataset_type, expected_result
+):
+    df = dataset_type.from_dict(data)
+    dataframe_operator = DataframeType({"value": df})
+    result = dataframe_operator.is_contained_by(
+        {"target": "target", "comparator": comparator}
+    )
+    assert result.equals(df.convert_to_series(expected_result))