BUG: df.sparse.to_coo() raising on duplicate colnames (#43491)

Author: mzeitlin11

asv_bench/benchmarks/sparse.py

@@ -91,6 +91,20 @@ def time_sparse_series_to_coo_single_level(self, sort_labels):
  self.ss_two_lvl.sparse.to_coo(sort_labels=sort_labels)
 
 
+class ToCooFrame:
+ def setup(self):
+ N = 10000
+ k = 10
+ arr = np.full((N, k), np.nan)
+ arr[0, 0] = 3.0
+ arr[12, 7] = -1.0
+ arr[0, 9] = 11.2
+ self.df = pd.DataFrame(arr, dtype=pd.SparseDtype("float"))
+
+ def time_to_coo(self):
+ self.df.sparse.to_coo()
+
+
 class Arithmetic:
 
  params = ([0.1, 0.01], [0, np.nan])

doc/source/whatsnew/v1.4.0.rst

@@ -431,6 +431,7 @@ Reshaping
 
 Sparse
 ^^^^^^
+- Bug in :meth:`DataFrame.sparse.to_coo` raising ``AttributeError`` when column names are not unique (:issue:`29564`)
 -
 -
 

pandas/core/arrays/sparse/accessor.py

@@ -339,12 +339,11 @@ def to_coo(self):
  dtype = dtype.subtype
 
  cols, rows, data = [], [], []
- for col, name in enumerate(self._parent):
- s = self._parent[name]
- row = s.array.sp_index.to_int_index().indices
+ for col, (_, ser) in enumerate(self._parent.iteritems()):
+ row = ser.array.sp_index.to_int_index().indices
  cols.append(np.repeat(col, len(row)))
  rows.append(row)
- data.append(s.array.sp_values.astype(dtype, copy=False))
+ data.append(ser.array.sp_values.astype(dtype, copy=False))
 
  cols = np.concatenate(cols)
  rows = np.concatenate(rows)

pandas/tests/arrays/sparse/test_accessor.py

@@ -71,7 +71,9 @@ def test_from_spmatrix_columns(self, columns):
  expected = pd.DataFrame(mat.toarray(), columns=columns).astype(dtype)
  tm.assert_frame_equal(result, expected)
 
- @pytest.mark.parametrize("colnames", [("A", "B"), (1, 2), (1, pd.NA), (0.1, 0.2)])
+ @pytest.mark.parametrize(
+ "colnames", [("A", "B"), (1, 2), (1, pd.NA), (0.1, 0.2), ("x", "x"), (0, 0)]
+ )
  @td.skip_if_no_scipy
  def test_to_coo(self, colnames):
  import scipy.sparse