[BUGFIX] droplabel when combined with other selectors does not exclude labels

.github/workflows/docs.yml

@@ -36,7 +36,7 @@ jobs:
       - name: Build docs
         run: mkdocs build
 
-      - uses: actions/upload-artifact@v3
+      - uses: actions/upload-artifact@v4
         with:
           name: website
           path: site/

janitor/functions/select.py

@@ -152,13 +152,13 @@ def select_columns(
 
         Exclude columns with the `DropLabel` class:
         >>> from janitor import DropLabel
-        >>> df.select_columns(DropLabel(slice("name", "awake")), "conservation")
-           brainwt   bodywt  conservation
-        0      NaN   50.000            lc
-        1  0.01550    0.480           NaN
-        2      NaN    1.350            nt
-        3  0.00029    0.019            lc
-        4  0.42300  600.000  domesticated
+        >>> df.select_columns(DropLabel(slice("name", "awake")))
+           brainwt   bodywt
+        0      NaN   50.000
+        1  0.01550    0.480
+        2      NaN    1.350
+        3  0.00029    0.019
+        4  0.42300  600.000
 
         Selection on MultiIndex columns:
         >>> d = {'num_legs': [4, 4, 2, 2],
@@ -673,7 +673,7 @@ def _index_dispatch(arg, df, axis):  # noqa: F811
 
     Returns an array of integers.
     """
-    level_label = {}
+
     index = getattr(df, axis)
     if not isinstance(index, pd.MultiIndex):
         return _select_index(list(arg), df, axis)
@@ -687,6 +687,7 @@ def _index_dispatch(arg, df, axis):  # noqa: F811
             "in the MultiIndex, and should either be all "
             "strings or integers."
         )
+    level_label = {}
     for key, value in arg.items():
         if isinstance(value, dispatch_callable):
             indexer = index.get_level_values(key)
@@ -757,14 +758,14 @@ def _index_dispatch(arg, df, axis):  # noqa: F811
 def _column_sel_dispatch(cols, df, axis):  # noqa: F811
     """
     Base function for selection on a Pandas Index object.
-    Returns the inverse of the passed label(s).
+    Returns the inverse of the passed label(s),
+    or the set difference if it is part of a list of labels.
 
     Returns an array of integers.
     """
     arr = _select_index(cols.label, df, axis)
     index = np.arange(getattr(df, axis).size)
-    arr = _index_converter(arr, index)
-    return np.delete(index, arr)
+    return _index_converter(arr, index)
 
 
 @_select_index.register(set)
@@ -797,27 +798,43 @@ def _index_dispatch(arg, df, axis):  # noqa: F811
             indices = index.get_indexer_for(list(arg))
             if (indices != -1).all():
                 return indices
-    # treat multiple DropLabel instances as a single unit
-    checks = (isinstance(entry, DropLabel) for entry in arg)
-    if sum(checks) > 1:
-        drop_labels = (entry for entry in arg if isinstance(entry, DropLabel))
-        drop_labels = [entry.label for entry in drop_labels]
-        drop_labels = DropLabel(drop_labels)
-        arg = [entry for entry in arg if not isinstance(entry, DropLabel)]
-        arg.append(drop_labels)
-    indices = [_select_index(entry, df, axis) for entry in arg]
+
+    include = []
+    exclude = []
+    for entry in arg:
+        if isinstance(entry, DropLabel):
+            exclude.append(entry)
+        else:
+            outcome = _select_index(entry, df, axis)
+            include.append(outcome)
+    if exclude:
+        if len(exclude) > 1:
+            exclude = [entry.label for entry in exclude]
+            exclude = DropLabel(exclude)
+        else:
+            exclude = exclude[0]
+        exclude = _select_index(exclude, df, axis)
+    len_exclude = len(exclude)
+    if len_exclude and not include:
+        index_arr = np.arange(getattr(df, axis).size)
+        return np.delete(index_arr, exclude)
+    if include and len_exclude:
+        include = [_index_converter(arr, index) for arr in include]
+        include = np.concatenate(include)
+        mask = np.isin(include, exclude)
+        return include[~mask]
     # single entry does not need to be combined
     # or materialized if possible;
     # this offers more performance
-    if len(indices) == 1:
-        if is_scalar(indices[0]):
-            return indices
-        indices = indices[0]
-        if is_list_like(indices):
-            indices = np.asanyarray(indices)
-        return indices
-    indices = [_index_converter(arr, index) for arr in indices]
-    return np.concatenate(indices)
+    if len(include) == 1:
+        if is_scalar(include[0]):
+            return include
+        include = include[0]
+        if is_list_like(include):
+            include = np.asanyarray(include)
+        return include
+    include = [_index_converter(arr, index) for arr in include]
+    return np.concatenate(include)
 
 
 def _index_converter(arr, index):

tests/functions/test_select_columns.py

@@ -44,6 +44,27 @@ def test_select_column_names_droplabel(dataframe, invert, expected):
     assert_frame_equal(df, dataframe[expected])
 
 
+def test_select_column_names_droplabel_mix():
+    "Test DropLabel, mixed with other labels"
+    data = {
+        "rowid": [1, 2],
+        "species": ["Adelie", "Adelie"],
+        "island": ["Torgersen", "Torgersen"],
+        "bill_length_mm": [39.1, 39.5],
+        "bill_depth_mm": [18.7, 17.4],
+        "flipper_length_mm": [181.0, 186.0],
+        "body_mass_g": [3750.0, 3800.0],
+        "sex": ["male", "female"],
+        "year": [2007, 2007],
+    }
+
+    df = pd.DataFrame(data)
+    selectors = [pd.api.types.is_numeric_dtype, DropLabel(["year", "rowid"])]
+    actual = df.select(selectors)
+    expected = df.select_dtypes("number").drop(columns=["year", "rowid"])
+    assert_frame_equal(actual, expected)
+
+
 @pytest.mark.functions
 def test_select_column_names_droplabel_multiple(dataframe):
     "Base DataFrame"