BUG: handling of array-like DataFrame elements in to_csv, make isnull work on lists of strings. close #1791

Author: wesm

RELEASE.rst

@@ -126,6 +126,8 @@ pandas 0.9.0
   - Prevent segfault due to MultiIndex not being supported in HDFStore table
     format (#1848)
   - Fix UnboundLocalError in Panel.__setitem__ and add better error (#1826)
+  - Fix to_csv issues with list of string entries. Isnull works on list of
+    strings now too (#1791)
 
 pandas 0.8.1
 ============

pandas/core/common.py

@@ -73,9 +73,13 @@ def _isnull_ndarraylike(obj):
     if values.dtype.kind in ('O', 'S'):
         # Working around NumPy ticket 1542
         shape = values.shape
-        result = np.empty(shape, dtype=bool)
-        vec = lib.isnullobj(values.ravel())
-        result[:] = vec.reshape(shape)
+
+        if values.dtype.kind == 'S':
+            result = np.zeros(values.shape, dtype=bool)
+        else:
+            result = np.empty(shape, dtype=bool)
+            vec = lib.isnullobj(values.ravel())
+            result[:] = vec.reshape(shape)
 
         if isinstance(obj, Series):
             result = Series(result, index=obj.index, copy=False)

pandas/core/frame.py

@@ -1088,7 +1088,7 @@ def _helper_csvexcel(self, writer, na_rep=None, cols=None,
                     row_fields = list(idx)
             for i, col in enumerate(cols):
                 val = series[col][j]
-                if isnull(val):
+                if lib.checknull(val):
                     val = na_rep
                 if float_format is not None and com.is_float(val):
                     val = float_format % val

pandas/tests/test_common.py

@@ -50,6 +50,10 @@ def test_isnull_lists():
     exp = np.array([[False], [False]])
     assert(np.array_equal(result, exp))
 
+    # list of strings
+    result = isnull(['foo', 'bar'])
+    assert(not result.any())
+
 def test_isnull_datetime():
     assert (not isnull(datetime.now()))
     assert notnull(datetime.now())

pandas/tests/test_series.py

@@ -2028,6 +2028,14 @@ def test_to_csv_float_format(self):
         assert_series_equal(rs, xp)
         os.remove(filename)
 
+    def test_to_csv_list_entries(self):
+        s = Series(['jack and jill','jesse and frank'])
+
+        split = s.str.split(r'\s+and\s+')
+
+        buf = StringIO()
+        split.to_csv(buf)
+
     def test_clip(self):
         val = self.ts.median()