BUG: float trunc in eval with py 2

Author: chris-b1

doc/source/whatsnew/v0.19.0.txt

@@ -1567,7 +1567,7 @@ Bug Fixes
 - Bug in ``DataFrame.to_csv()`` with ``MultiIndex`` columns in which a stray empty line was added (:issue:`6618`)
 - Bug in ``DatetimeIndex``, ``TimedeltaIndex`` and ``PeriodIndex.equals()`` may return ``True`` when input isn't ``Index`` but contains the same values (:issue:`13107`)
 - Bug in assignment against datetime with timezone may not work if it contains datetime near DST boundary (:issue:`14146`)
-
+- Bug in ``pd.eval()`` truncating long float literals with python 2 (:issue:`14241`)
 - Bug in ``Index`` raises ``KeyError`` displaying incorrect column when column is not in the df and columns contains duplicate values (:issue:`13822`)
 - Bug in ``Period`` and ``PeriodIndex`` creating wrong dates when frequency has combined offset aliases (:issue:`13874`)
 - Bug in ``.to_string()`` when called with an integer ``line_width`` and ``index=False`` raises an UnboundLocalError exception because ``idx`` referenced before assignment.

pandas/computation/ops.py

@@ -166,6 +166,11 @@ def _resolve_name(self):
     def name(self):
         return self.value
 
+    def __unicode__(self):
+        # in python 2 str() of float
+        # can truncate shorter than repr()
+        return repr(self.name)
+
 
 _bool_op_map = {'not': '~', 'and': '&', 'or': '|'}
 

pandas/computation/tests/test_eval.py

@@ -678,6 +678,18 @@ def test_line_continuation(self):
         result = pd.eval(exp, engine=self.engine, parser=self.parser)
         self.assertEqual(result, 12)
 
+    def test_float_truncation(self):
+        # GH 14241
+        exp = '1000000000.006'
+        result = pd.eval(exp, engine=self.engine, parser=self.parser)
+        expected = np.float64(exp)
+        self.assertEqual(result, expected)
+
+        df = pd.DataFrame([{"A": 1000000000.0099}])
+        cutoff = 1000000000.006
+        result = df.query("A < %.3f" % cutoff)
+        self.assertTrue(result.empty)
+
 
 class TestEvalNumexprPython(TestEvalNumexprPandas):