Update expected behavior of groupby.idxmax with strings

Previously, Series.idxmax with string data raised a ValueError, due to
an attempt to coerce the strings to floats in the `nanops` module.
DataFrame.groupby.idxmax would consequently drop columns containing
string data.

Author: DGrady

pandas/tests/groupby/test_groupby.py

@@ -2339,7 +2339,7 @@ def test_non_cython_api(self):
         assert_frame_equal(result, expected)
 
         # idxmax
-        expected = DataFrame([[0], [nan]], columns=['B'], index=[1, 3])
+        expected = DataFrame([[0.0, 0.0], [nan, 2.0]], columns=['B', 'C'], index=[1, 3])
         expected.index.name = 'A'
         result = g.idxmax()
         assert_frame_equal(result, expected)