BUG: Work-around a numpy regression

numpy 1.7.0 erroneously raises IndexError instead of ValueError
from ndarray.item() when the array is not of length 1. This can be
seen as a failure of

    computation.tests.test_eval.TestScope.test_global_scope

for the cases that engine='numexpr'.

Absorb the splatter from this regression by explicitly catching the
erroneous IndexError.

Author: neirbowj

pandas/computation/align.py

@@ -251,6 +251,9 @@ def _reconstruct_object(typ, obj, axes, dtype):
 
     try:
         ret = ret_value.item()
-    except ValueError:
+    except (ValueError, IndexError):
+        # XXX: we catch IndexError to absorb a
+        # regression in numpy 1.7.0
+        # fixed by numpy/numpy@04b89c63
         ret = ret_value
     return ret