BUG: Properly validate and parse nrows in read_csv

1) Allows `float` values for `nrows` for the Python engine  2)
Prevents abuse of the `nrows` argument for the CParser (e.g. you can
passing `nrows=1.2`)    Closes #10476.

Author: gfyoung <[email protected]>

Closes #13275 from gfyoung/nrows-bug-validate and squashes the following commits:

d856051 [gfyoung] BUG: Properly validate and parse nrows in read_csv

Author: gfyoung

doc/source/whatsnew/v0.18.2.txt

@@ -252,3 +252,4 @@ Bug Fixes
 - Bug in ``groupby`` where ``apply`` returns different result depending on whether first result is ``None`` or not (:issue:`12824`)
 
 - Bug in ``Categorical.remove_unused_categories()`` changes ``.codes`` dtype to platform int (:issue:`13261`)
+- Bug in ``pd.read_csv`` in which the ``nrows`` argument was not properly validated for both engines (:issue:`10476`)

pandas/io/parsers.py

@@ -272,6 +272,26 @@
 """ % (_parser_params % (_fwf_widths, ''))
 
 
+def _validate_nrows(nrows):
+    """
+    Checks whether the 'nrows' parameter for parsing is either
+    an integer OR float that can SAFELY be cast to an integer
+    without losing accuracy. Raises a ValueError if that is
+    not the case.
+    """
+    msg = "'nrows' must be an integer"
+
+    if nrows is not None:
+        if com.is_float(nrows):
+            if int(nrows) != nrows:
+                raise ValueError(msg)
+            nrows = int(nrows)
+        elif not com.is_integer(nrows):
+            raise ValueError(msg)
+
+    return nrows
+
+
 def _read(filepath_or_buffer, kwds):
     "Generic reader of line files."
     encoding = kwds.get('encoding', None)
@@ -311,14 +331,14 @@ def _read(filepath_or_buffer, kwds):
 
     # Extract some of the arguments (pass chunksize on).
     iterator = kwds.get('iterator', False)
-    nrows = kwds.pop('nrows', None)
     chunksize = kwds.get('chunksize', None)
+    nrows = _validate_nrows(kwds.pop('nrows', None))
 
     # Create the parser.
     parser = TextFileReader(filepath_or_buffer, **kwds)
 
     if (nrows is not None) and (chunksize is not None):
-        raise NotImplementedError("'nrows' and 'chunksize' can not be used"
+        raise NotImplementedError("'nrows' and 'chunksize' cannot be used"
                                   " together yet.")
     elif nrows is not None:
         return parser.read(nrows)

pandas/io/tests/parser/common.py

@@ -391,10 +391,23 @@ def test_int_conversion(self):
         self.assertEqual(data['B'].dtype, np.int64)
 
     def test_read_nrows(self):
-        df = self.read_csv(StringIO(self.data1), nrows=3)
         expected = self.read_csv(StringIO(self.data1))[:3]
+
+        df = self.read_csv(StringIO(self.data1), nrows=3)
         tm.assert_frame_equal(df, expected)
 
+        # see gh-10476
+        df = self.read_csv(StringIO(self.data1), nrows=3.0)
+        tm.assert_frame_equal(df, expected)
+
+        msg = "must be an integer"
+
+        with tm.assertRaisesRegexp(ValueError, msg):
+            self.read_csv(StringIO(self.data1), nrows=1.2)
+
+        with tm.assertRaisesRegexp(ValueError, msg):
+            self.read_csv(StringIO(self.data1), nrows='foo')
+
     def test_read_chunksize(self):
         reader = self.read_csv(StringIO(self.data1), index_col=0, chunksize=2)
         df = self.read_csv(StringIO(self.data1), index_col=0)
@@ -815,11 +828,6 @@ def test_ignore_leading_whitespace(self):
         expected = DataFrame({'a': [1, 4, 7], 'b': [2, 5, 8], 'c': [3, 6, 9]})
         tm.assert_frame_equal(result, expected)
 
-    def test_nrows_and_chunksize_raises_notimplemented(self):
-        data = 'a b c'
-        self.assertRaises(NotImplementedError, self.read_csv, StringIO(data),
-                          nrows=10, chunksize=5)
-
     def test_chunk_begins_with_newline_whitespace(self):
         # see gh-10022
         data = '\n hello\nworld\n'

pandas/io/tests/parser/test_unsupported.py

@@ -30,6 +30,15 @@ def test_mangle_dupe_cols_false(self):
                 read_csv(StringIO(data), engine=engine,
                          mangle_dupe_cols=False)
 
+    def test_nrows_and_chunksize(self):
+        data = 'a b c'
+        msg = "cannot be used together yet"
+
+        for engine in ('c', 'python'):
+            with tm.assertRaisesRegexp(NotImplementedError, msg):
+                read_csv(StringIO(data), engine=engine,
+                         nrows=10, chunksize=5)
+
     def test_c_engine(self):
         # see gh-6607
         data = 'a b c\n1 2 3'