remove incomplete braid solution

Author: JohnMount

data_algebra/solutions.py

@@ -341,111 +341,6 @@ def last_observed_carried_forward(
     return ops
 
 
-def braid_data(
-    *,
-    d_state: ViewRepresentation,
-    d_event: ViewRepresentation,
-    order_by: Iterable[str],
-    partition_by: Optional[Iterable[str]] = None,
-    state_value_column_name: str,
-    event_value_column_names: Iterable[str],
-    source_id_column: str = "record_type",
-    state_row_mark: str = "state_row",
-    event_row_mark: str = "event_row",
-    stand_in_values: Dict,
-    locf_to_use_column_name: str = "locf_to_use",
-    locf_non_null_rank_column_name: str = "locf_non_null_rank",
-    locf_tiebreaker_column_name: str = "locf_tiebreaker",
-) -> ViewRepresentation:
-    """
-    Mix data from two sources, ordering by order_by columns and carrying forward observations
-    on d_state value column.
-
-    :param d_state: ViewRepresentation representation of state by order_by.
-    :param d_event: ViewRepresentation representation of events by order_by.
-    :param order_by: columns to order by (non empty list of column names)
-    :param partition_by: optional partitioning column names
-    :param state_value_column_name: column to copy from d_state and propagate forward
-    :param event_value_column_names: columns to copy from d_event
-    :param source_id_column: name for source identification column.
-    :param state_row_mark: source annotation of state rows.
-    :param event_row_mark: source annotation of event rows.
-    :param stand_in_values: dictionary stand in values to use for state_value_column_name and event_value_column_names
-            needed to get column types correct, replaced by None and not passed further.
-    :param locf_to_use_column_name: name for a temporary values column
-    :param locf_non_null_rank_column_name: name for a temporary values column
-    :param locf_tiebreaker_column_name: name for a temporary values column
-    :return: ops
-    """
-    assert isinstance(d_state, ViewRepresentation)
-    assert isinstance(d_event, ViewRepresentation)
-    assert not isinstance(order_by, str)
-    order_by = list(order_by)
-    assert len(order_by) > 0
-    if partition_by is not None:
-        assert not isinstance(partition_by, str)
-        partition_by = list(partition_by)
-    else:
-        partition_by = []
-    assert isinstance(state_value_column_name, str)
-    assert not isinstance(event_value_column_names, str)
-    event_value_column_names = list(event_value_column_names)
-    assert isinstance(source_id_column, str)
-    assert isinstance(state_row_mark, str)
-    assert isinstance(event_row_mark, str)
-    assert isinstance(locf_to_use_column_name, str)
-    assert isinstance(locf_non_null_rank_column_name, str)
-    assert isinstance(locf_tiebreaker_column_name, str)
-    assert isinstance(stand_in_values, dict)
-    together = (
-        d_state.extend({k: stand_in_values[k] for k in event_value_column_names})
-        .select_columns(
-            partition_by
-            + order_by
-            + [state_value_column_name]
-            + event_value_column_names
-        )
-        .concat_rows(
-            b=(
-                d_event.extend(
-                    {state_value_column_name: stand_in_values[state_value_column_name]}
-                ).select_columns(
-                    partition_by
-                    + order_by
-                    + [state_value_column_name]
-                    + event_value_column_names
-                )
-            ),
-            id_column=source_id_column,
-            a_name=state_row_mark,
-            b_name=event_row_mark,
-        )
-        # clear out stand-in values
-        .extend(
-            {
-                state_value_column_name: f'({source_id_column} == "{event_row_mark}").if_else(None, {state_value_column_name})'
-            }
-        )
-        .extend(
-            {
-                k: f'({source_id_column} == "{state_row_mark}").if_else(None, {k})'
-                for k in event_value_column_names
-            }
-        )
-    )
-    ops = last_observed_carried_forward(
-        together,
-        order_by=order_by,
-        partition_by=partition_by,
-        value_column_name=state_value_column_name,
-        selection_predicate="is_null()",
-        locf_to_use_column_name=locf_to_use_column_name,
-        locf_non_null_rank_column_name=locf_non_null_rank_column_name,
-        locf_tiebreaker_column_name=locf_tiebreaker_column_name,
-    )
-    return ops
-
-
 def rank_to_average(
     d: ViewRepresentation,
     *,

tests/test_braid.py

@@ -68,9 +68,10 @@ def test_idiom_extend_special_count():
     expect = data_algebra.data_model.default_data_model().pd.DataFrame({"count": [4]})
 
     with pytest.warns(UserWarning):
+        # warning is db adapter saying to not use this fn
         data_algebra.test_util.check_transform(
             ops=ops, data=d, expect=expect, empty_produces_empty=False,
-            try_on_Polars=False,  # TODO: turn this on
+            try_on_Polars=False,  # TODO: turn this on, don't use cumsum() in project on Polars
         )