BUG: incorrect handling of scipy.sparse.dok formats (#16197)

Author: kristopheryahoo

doc/source/whatsnew/v0.20.1.txt

@@ -66,8 +66,7 @@ Groupby/Resample/Rolling
 Sparse
 ^^^^^^
 
-
-
+- Bug in construction of SparseDataFrame from ``scipy.sparse.dok_matrix`` (:issue:`16179`)
 
 Reshaping
 ^^^^^^^^^

pandas/core/sparse/frame.py

@@ -190,8 +190,8 @@ def _init_spmatrix(self, data, index, columns, dtype=None,
         values = Series(data.data, index=data.row, copy=False)
         for col, rowvals in values.groupby(data.col):
             # get_blocks expects int32 row indices in sorted order
+            rowvals = rowvals.sort_index()
             rows = rowvals.index.values.astype(np.int32)
-            rows.sort()
             blocs, blens = get_blocks(rows)
 
             sdict[columns[col]] = SparseSeries(

pandas/tests/sparse/test_frame.py

@@ -1146,8 +1146,8 @@ def test_isnotnull(self):
         tm.assert_frame_equal(res.to_dense(), exp)
 
 
-@pytest.mark.parametrize('index', [None, list('ab')])  # noqa: F811
-@pytest.mark.parametrize('columns', [None, list('cd')])
+@pytest.mark.parametrize('index', [None, list('abc')])  # noqa: F811
+@pytest.mark.parametrize('columns', [None, list('def')])
 @pytest.mark.parametrize('fill_value', [None, 0, np.nan])
 @pytest.mark.parametrize('dtype', [bool, int, float, np.uint16])
 def test_from_to_scipy(spmatrix, index, columns, fill_value, dtype):
@@ -1156,7 +1156,8 @@ def test_from_to_scipy(spmatrix, index, columns, fill_value, dtype):
 
     # Make one ndarray and from it one sparse matrix, both to be used for
     # constructing frames and comparing results
-    arr = np.eye(2, dtype=dtype)
+    arr = np.eye(3, dtype=dtype)
+    arr[0, 1] = dtype(2)
     try:
         spm = spmatrix(arr)
         assert spm.dtype == arr.dtype