More careful change

Author: charlesdong1991

pandas/plotting/_matplotlib/boxplot.py

@@ -315,21 +315,20 @@ def plot_group(keys, values, ax):
  ax = plt.gca()
  data = data._get_numeric_data()
 
- # if columns is None, use all numeric columns of data; if data columns
- # is MultiIndex, which means a Groupby has been applied before, select
- # data using new grouped column names; if data columns is Index, select
- # data simply using columns
+ # if columns is None, use all numeric columns of data, so directly pass
+ # if the given 'column' are subset of column index, no matter if data column
+ # is multiindex or index, get the subset of data directly
+ # if given 'column' is not subset, and data columns is multiindex, and then
+ # query the columns if column contains at least one element from 'columns'
  if columns is None:
- columns = data.columns
+ pass
+ elif set(column).issubset(data.columns):
+ data = data.loc[:, columns]
  elif isinstance(data.columns, pd.MultiIndex):
-
- # reselect columns with after-groupby multi-index columns
- data = data.loc[:, pd.IndexSlice[:, columns]]
- columns = data.columns
- elif isinstance(data.columns, pd.Index):
+ columns = [col for col in data.columns if set(columns).intersection(col)]
  data = data.loc[:, columns]
- columns = data.columns
 
+ columns = data.columns
  result = plot_group(columns, data.values.T, ax)
  ax.grid(grid)
 

pandas/tests/plotting/test_boxplot_method.py

@@ -455,3 +455,28 @@ def test_groupby_boxplot_subplots_false(self, col, expected_xticklabel):
  # check if xticks labels are plotted correctly
  result_xticklabel = [x.get_text() for x in axes.get_xticklabels()]
  assert expected_xticklabel == result_xticklabel
+
+ @pytest.mark.parametrize(
+ "col, expected_xticklabel",
+ [
+ ([("bar", "one"), ("bar", "two")], ["(bar, one)", "(bar, two)"]),
+ ("bar", ["(bar, one)", "(bar, two)"]),
+ (["two"], ["(bar, two)", "(baz, two)", "(foo, two)", "(qux, two)"]),
+ ],
+ )
+ def test_boxplot_multiindex_column(self, col, expected_xticklabel):
+ # this is test the boxplot on multi-index column cases
+ arrays = [
+ ["bar", "bar", "baz", "baz", "foo", "foo", "qux", "qux"],
+ ["one", "two", "one", "two", "one", "two", "one", "two"],
+ ]
+ tuples = list(zip(*arrays))
+ index = MultiIndex.from_tuples(tuples, names=["first", "second"])
+ df = DataFrame(np.random.randn(3, 8), index=["A", "B", "C"], columns=index)
+
+ # check if df.boxplot works
+ axes = _check_plot_works(df.boxplot, column=col, return_type="axes")
+
+ # check if xticks labels are plotted correctly
+ result_xticklabel = [x.get_text() for x in axes.get_xticklabels()]
+ assert expected_xticklabel == result_xticklabel