Add more specific error message when user passes incorrect matrix format to from_coo

doc/source/whatsnew/v0.25.0.rst

@@ -692,7 +692,7 @@ Sparse
 - Significant speedup in :class:`SparseArray` initialization that benefits most operations, fixing performance regression introduced in v0.20.0 (:issue:`24985`)
 - Bug in :class:`SparseFrame` constructor where passing ``None`` as the data would cause ``default_fill_value`` to be ignored (:issue:`16807`)
 - Bug in :class:`SparseDataFrame` when adding a column in which the length of values does not match length of index, ``AssertionError`` is raised instead of raising ``ValueError`` (:issue:`25484`)
-
+- Introduce a better error message in :meth:`Series.sparse.from_coo` so it returns a ``TypeError`` for inputs that are not coo matrices (:issue:`26554`)
 
 Other
 ^^^^^

pandas/core/sparse/scipy_sparse.py

@@ -130,10 +130,19 @@ def _coo_to_sparse_series(A, dense_index: bool = False,
     Returns
     -------
     Series or SparseSeries
+
+    Raises
+    ------
+    TypeError if A is not a coo_matrix
+
     """
     from pandas import SparseDtype
 
-    s = Series(A.data, MultiIndex.from_arrays((A.row, A.col)))
+    try:
+        s = Series(A.data, MultiIndex.from_arrays((A.row, A.col)))
+    except AttributeError:
+        raise TypeError('Expected coo_matrix. Got {} instead.'
+                        .format(type(A).__name__))
     s = s.sort_index()
     if sparse_series:
         # TODO(SparseSeries): remove this and the sparse_series keyword.

pandas/tests/arrays/sparse/test_accessor.py

@@ -119,3 +119,13 @@ def test_series_from_coo(self, dtype, dense_index):
             )
 
         tm.assert_series_equal(result, expected)
+
+    @td.skip_if_no_scipy
+    def test_series_from_coo_incorrect_format_raises(self):
+        # gh-26554
+        import scipy.sparse
+        m = scipy.sparse.csr_matrix(np.array([[0, 1], [0, 0]]))
+        with pytest.raises(TypeError,
+                           match='Expected coo_matrix. Got csr_matrix instead.'
+                           ):
+            pd.Series.sparse.from_coo(m)