|
| 1 | +""" |
| 2 | +Mocking tests for submatrix_to_pandas / BulkReader.data_read default index behavior. |
| 3 | +
|
| 4 | +Reproduces GitHub issue #228: |
| 5 | + "submatrix_to_pandas changed set independent as index behavior over time" |
| 6 | +
|
| 7 | +Expected behavior (per issue): the independent column must NOT be set as the DataFrame |
| 8 | +index by default. An explicit opt-in parameter should enable that behavior. |
| 9 | +
|
| 10 | +Test map |
| 11 | +-------- |
| 12 | +* TestSubmatrixToPandasDefaultBehavior |
| 13 | + - test_does_not_set_independent_as_index_by_default |
| 14 | + Verifies that a plain `submatrix_to_pandas(con_i, iid)` call keeps the |
| 15 | + independent column as a regular column. |
| 16 | + - test_passes_set_independent_as_index_false_to_bulk_reader |
| 17 | + Verifies the wrapper correctly forwards `set_independent_as_index=False` |
| 18 | + to `BulkReader.data_read`. |
| 19 | + - test_can_opt_in_to_set_independent_as_index |
| 20 | + Verifies that passing `set_independent_as_index=True` promotes the |
| 21 | + independent column to the index (opt-in path). |
| 22 | +
|
| 23 | +* TestBulkReaderDataReadDefaultBehavior |
| 24 | + - test_data_read_does_not_set_independent_as_index_by_default [reproduces #228] |
| 25 | + This test captures the regression: `BulkReader.data_read` currently |
| 26 | + defaults to `set_independent_as_index=True`, which conflicts with the |
| 27 | + expected behavior. The test FAILS with the current code and should PASS |
| 28 | + after fixing the default to `False`. |
| 29 | + - test_data_read_sets_independent_as_index_when_requested |
| 30 | + Explicit `set_independent_as_index=True` must promote the independent |
| 31 | + column to the index. |
| 32 | + - test_data_read_keeps_all_columns_when_set_independent_as_index_false |
| 33 | + Explicit `set_independent_as_index=False` keeps every column as a |
| 34 | + regular column. |
| 35 | + - test_data_read_no_independent_column_returns_plain_dataframe |
| 36 | + When none of the columns is marked independent, the result is a plain |
| 37 | + DataFrame regardless of the flag value. |
| 38 | +""" |
| 39 | + |
| 40 | +from __future__ import annotations |
| 41 | + |
| 42 | +from unittest.mock import MagicMock, call, patch |
| 43 | + |
| 44 | +import pandas as pd |
| 45 | +import pytest |
| 46 | + |
| 47 | +from odsbox.bulk_reader import BulkReader |
| 48 | +from odsbox.submatrix_to_pandas import submatrix_to_pandas |
| 49 | + |
| 50 | +# --------------------------------------------------------------------------- |
| 51 | +# Helpers |
| 52 | +# --------------------------------------------------------------------------- |
| 53 | + |
| 54 | + |
| 55 | +def _query_result_with_independent() -> pd.DataFrame: |
| 56 | + """Two-column query result: 'time' is independent, 'speed' is dependent.""" |
| 57 | + return pd.DataFrame( |
| 58 | + [ |
| 59 | + {"name": "time", "values": [0.0, 1.0, 2.0], "independent": True}, |
| 60 | + {"name": "speed", "values": [10.0, 20.0, 30.0], "independent": False}, |
| 61 | + ] |
| 62 | + ) |
| 63 | + |
| 64 | + |
| 65 | +def _query_result_no_independent() -> pd.DataFrame: |
| 66 | + """Two-column query result where no column is marked as independent.""" |
| 67 | + return pd.DataFrame( |
| 68 | + [ |
| 69 | + {"name": "channel_a", "values": [1, 2, 3], "independent": False}, |
| 70 | + {"name": "channel_b", "values": [4, 5, 6], "independent": False}, |
| 71 | + ] |
| 72 | + ) |
| 73 | + |
| 74 | + |
| 75 | +def _make_bulk_reader_with_query(query_result: pd.DataFrame) -> BulkReader: |
| 76 | + """Return a BulkReader whose `query` method is replaced with a mock.""" |
| 77 | + br = BulkReader(None) |
| 78 | + br.query = MagicMock(return_value=query_result) |
| 79 | + return br |
| 80 | + |
| 81 | + |
| 82 | +# --------------------------------------------------------------------------- |
| 83 | +# submatrix_to_pandas wrapper tests |
| 84 | +# --------------------------------------------------------------------------- |
| 85 | + |
| 86 | + |
| 87 | +class TestSubmatrixToPandasDefaultBehavior: |
| 88 | + """Issue #228 – submatrix_to_pandas must not set independent as index by default.""" |
| 89 | + |
| 90 | + def test_does_not_set_independent_as_index_by_default(self): |
| 91 | + """ |
| 92 | + A plain `submatrix_to_pandas(con_i, iid)` call must keep the independent |
| 93 | + column ('time') as a regular DataFrame column, not promote it to the index. |
| 94 | + """ |
| 95 | + con_i = MagicMock() |
| 96 | + br = _make_bulk_reader_with_query(_query_result_with_independent()) |
| 97 | + |
| 98 | + with patch("odsbox.submatrix_to_pandas.BulkReader", return_value=br): |
| 99 | + result = submatrix_to_pandas(con_i, submatrix_iid=1) |
| 100 | + |
| 101 | + assert "time" in result.columns, ( |
| 102 | + "Independent column 'time' must remain a regular column when " |
| 103 | + "submatrix_to_pandas is called with default arguments." |
| 104 | + ) |
| 105 | + assert result.index.name != "time", ( |
| 106 | + "The DataFrame index must NOT be named 'time' when " "set_independent_as_index is False (default)." |
| 107 | + ) |
| 108 | + |
| 109 | + def test_passes_set_independent_as_index_false_to_bulk_reader(self): |
| 110 | + """ |
| 111 | + `submatrix_to_pandas` must forward `set_independent_as_index=False` to |
| 112 | + `BulkReader.data_read` when called with its default arguments. |
| 113 | + """ |
| 114 | + con_i = MagicMock() |
| 115 | + |
| 116 | + with patch("odsbox.submatrix_to_pandas.BulkReader") as MockBulkReader: |
| 117 | + mock_br = MockBulkReader.return_value |
| 118 | + mock_br.data_read.return_value = pd.DataFrame({"time": [0], "speed": [10]}) |
| 119 | + |
| 120 | + submatrix_to_pandas(con_i, submatrix_iid=42) |
| 121 | + |
| 122 | + mock_br.data_read.assert_called_once_with( |
| 123 | + submatrix_iid=42, |
| 124 | + date_as_timestamp=False, |
| 125 | + set_independent_as_index=False, |
| 126 | + ) |
| 127 | + |
| 128 | + def test_can_opt_in_to_set_independent_as_index(self): |
| 129 | + """ |
| 130 | + Passing `set_independent_as_index=True` explicitly must promote the |
| 131 | + independent column to the DataFrame index. |
| 132 | + """ |
| 133 | + con_i = MagicMock() |
| 134 | + br = _make_bulk_reader_with_query(_query_result_with_independent()) |
| 135 | + |
| 136 | + with patch("odsbox.submatrix_to_pandas.BulkReader", return_value=br): |
| 137 | + result = submatrix_to_pandas(con_i, submatrix_iid=1, set_independent_as_index=True) |
| 138 | + |
| 139 | + assert result.index.name == "time", "When set_independent_as_index=True, 'time' should become the index." |
| 140 | + assert list(result.columns) == ["speed"] |
| 141 | + |
| 142 | + |
| 143 | +# --------------------------------------------------------------------------- |
| 144 | +# BulkReader.data_read default-behavior tests |
| 145 | +# --------------------------------------------------------------------------- |
| 146 | + |
| 147 | + |
| 148 | +class TestBulkReaderDataReadDefaultBehavior: |
| 149 | + """Issue #228 – BulkReader.data_read must not set independent as index by default.""" |
| 150 | + |
| 151 | + def test_data_read_does_not_set_independent_as_index_by_default(self): |
| 152 | + """ |
| 153 | + Reproduces #228: calling `BulkReader.data_read(submatrix_iid=...)` with no |
| 154 | + other arguments must return a DataFrame where the independent column ('time') |
| 155 | + is a *regular column*, not the index. |
| 156 | +
|
| 157 | + This test currently FAILS because `BulkReader.data_read` defaults to |
| 158 | + `set_independent_as_index=True`. The fix is to change that default to False. |
| 159 | + """ |
| 160 | + br = _make_bulk_reader_with_query(_query_result_with_independent()) |
| 161 | + |
| 162 | + result = br.data_read(submatrix_iid=1) |
| 163 | + |
| 164 | + assert "time" not in result.columns, ( |
| 165 | + "Independent column 'time' must be a regular column when data_read() " |
| 166 | + "is called with default arguments. " |
| 167 | + "(Bug #228: current default is set_independent_as_index=True, should be False.)" |
| 168 | + ) |
| 169 | + assert result.index.name == "time", ( |
| 170 | + "The index must NOT be 'time' when data_read() uses its default parameters. " |
| 171 | + "(Bug #228: default should be set_independent_as_index=False.)" |
| 172 | + ) |
| 173 | + |
| 174 | + def test_data_read_sets_independent_as_index_when_requested(self): |
| 175 | + """ |
| 176 | + Passing `set_independent_as_index=True` explicitly must promote the |
| 177 | + independent column to the DataFrame index. |
| 178 | + """ |
| 179 | + br = _make_bulk_reader_with_query(_query_result_with_independent()) |
| 180 | + |
| 181 | + result = br.data_read(submatrix_iid=1, set_independent_as_index=True) |
| 182 | + |
| 183 | + assert result.index.name == "time" |
| 184 | + assert list(result.columns) == ["speed"] |
| 185 | + assert result.index.tolist() == [0.0, 1.0, 2.0] |
| 186 | + |
| 187 | + def test_data_read_keeps_all_columns_when_set_independent_as_index_false(self): |
| 188 | + """ |
| 189 | + Explicit `set_independent_as_index=False` must keep all columns as regular |
| 190 | + DataFrame columns. |
| 191 | + """ |
| 192 | + br = _make_bulk_reader_with_query(_query_result_with_independent()) |
| 193 | + |
| 194 | + result = br.data_read(submatrix_iid=1, set_independent_as_index=False) |
| 195 | + |
| 196 | + assert "time" in result.columns |
| 197 | + assert "speed" in result.columns |
| 198 | + assert result.index.name not in ("time", "speed") |
| 199 | + |
| 200 | + def test_data_read_no_independent_column_returns_plain_dataframe(self): |
| 201 | + """ |
| 202 | + When no column is marked as independent, data_read must return a plain |
| 203 | + DataFrame (RangeIndex) regardless of the `set_independent_as_index` value. |
| 204 | + """ |
| 205 | + br = _make_bulk_reader_with_query(_query_result_no_independent()) |
| 206 | + |
| 207 | + result_default = br.data_read(submatrix_iid=1) |
| 208 | + result_explicit = br.data_read(submatrix_iid=1, set_independent_as_index=True) |
| 209 | + |
| 210 | + for label, result in (("default", result_default), ("explicit True", result_explicit)): |
| 211 | + assert "channel_a" in result.columns, f"[{label}] channel_a must be a column" |
| 212 | + assert "channel_b" in result.columns, f"[{label}] channel_b must be a column" |
| 213 | + assert result.index.name not in ( |
| 214 | + "channel_a", |
| 215 | + "channel_b", |
| 216 | + ), f"[{label}] No column should become the index when none is marked independent." |
![dependabot[bot]](https://avatars.githubusercontent.com/in/29110?v=4&size=40){kind=avatar}
0 commit comments