BUG: Patch read_csv NA values behaviour

Patches the following behaviour when `na_values` is passed in as a
dictionary:    1. Prevent aliasing in case `na_values` was defined in
a broader scope.  2. Respect column indices as keys when doing NA
conversions.    Closes #14203.

Author: gfyoung <[email protected]>

Closes #14751 from gfyoung/csv-na-values-patching and squashes the following commits:

cac422c [gfyoung] BUG: Respect column indices for dict-like na_values
1439c27 [gfyoung] BUG: Prevent aliasing of dict na_values

Author: gfyoung

doc/source/whatsnew/v0.19.2.txt

@@ -40,6 +40,8 @@ Bug Fixes
 
 - Compat with ``dateutil==2.6.0``; segfault reported in the testing suite (:issue:`14621`)
 - Allow ``nanoseconds`` in ``Timestamp.replace`` as a kwarg (:issue:`14621`)
+- Bug in ``pd.read_csv`` in which aliasing was being done for ``na_values`` when passed in as a dictionary (:issue:`14203`)
+- Bug in ``pd.read_csv`` in which column indices for a dict-like ``na_values`` were not being respected (:issue:`14203`)
 - Bug in ``pd.read_csv`` where reading files fails, if the number of headers is equal to the number of lines in the file (:issue:`14515`)
 - Bug in ``pd.read_csv`` for the Python engine in which an unhelpful error message was being raised when multi-char delimiters were not being respected with quotes (:issue:`14582`)
 - Fix bugs (:issue:`14734`, :issue:`13654`) in ``pd.read_sas`` and ``pandas.io.sas.sas7bdat.SAS7BDATReader`` that caused problems when reading a SAS file incrementally.

pandas/io/parsers.py

@@ -2055,9 +2055,27 @@ def _clean_mapping(mapping):
         else:
             clean_dtypes = _clean_mapping(self.dtype)
 
-        return self._convert_to_ndarrays(data, self.na_values, self.na_fvalues,
-                                         self.verbose, clean_conv,
-                                         clean_dtypes)
+        # Apply NA values.
+        clean_na_values = {}
+        clean_na_fvalues = {}
+
+        if isinstance(self.na_values, dict):
+            for col in self.na_values:
+                na_value = self.na_values[col]
+                na_fvalue = self.na_fvalues[col]
+
+                if isinstance(col, int) and col not in self.orig_names:
+                    col = self.orig_names[col]
+
+                clean_na_values[col] = na_value
+                clean_na_fvalues[col] = na_fvalue
+        else:
+            clean_na_values = self.na_values
+            clean_na_fvalues = self.na_fvalues
+
+        return self._convert_to_ndarrays(data, clean_na_values,
+                                         clean_na_fvalues, self.verbose,
+                                         clean_conv, clean_dtypes)
 
     def _to_recarray(self, data, columns):
         dtypes = []
@@ -2767,6 +2785,7 @@ def _clean_na_values(na_values, keep_default_na=True):
             na_values = []
         na_fvalues = set()
     elif isinstance(na_values, dict):
+        na_values = na_values.copy()  # Prevent aliasing.
         if keep_default_na:
             for k, v in compat.iteritems(na_values):
                 if not is_list_like(v):

pandas/io/tests/parser/na_values.py

@@ -266,3 +266,26 @@ def test_na_values_scalar(self):
         out = self.read_csv(StringIO(data), names=names,
                             na_values={'a': 2, 'b': 1})
         tm.assert_frame_equal(out, expected)
+
+    def test_na_values_dict_aliasing(self):
+        na_values = {'a': 2, 'b': 1}
+        na_values_copy = na_values.copy()
+
+        names = ['a', 'b']
+        data = '1,2\n2,1'
+
+        expected = DataFrame([[1.0, 2.0], [np.nan, np.nan]], columns=names)
+        out = self.read_csv(StringIO(data), names=names, na_values=na_values)
+
+        tm.assert_frame_equal(out, expected)
+        tm.assert_dict_equal(na_values, na_values_copy)
+
+    def test_na_values_dict_col_index(self):
+        # see gh-14203
+
+        data = 'a\nfoo\n1'
+        na_values = {0: 'foo'}
+
+        out = self.read_csv(StringIO(data), na_values=na_values)
+        expected = DataFrame({'a': [np.nan, 1]})
+        tm.assert_frame_equal(out, expected)

pandas/parser.pyx

@@ -1262,19 +1262,23 @@ cdef class TextReader:
             return None, set()
 
         if isinstance(self.na_values, dict):
+            key = None
             values = None
+
             if name is not None and name in self.na_values:
-                values = self.na_values[name]
-                if values is not None and not isinstance(values, list):
-                    values = list(values)
-                fvalues = self.na_fvalues[name]
-                if fvalues is not None and not isinstance(fvalues, set):
-                    fvalues = set(fvalues)
-            else:
-                if i in self.na_values:
-                    return self.na_values[i], self.na_fvalues[i]
-                else:
-                    return _NA_VALUES, set()
+                key = name
+            elif i in self.na_values:
+                key = i
+            else:  # No na_values provided for this column.
+                return _NA_VALUES, set()
+
+            values = self.na_values[key]
+            if values is not None and not isinstance(values, list):
+                values = list(values)
+
+            fvalues = self.na_fvalues[key]
+            if fvalues is not None and not isinstance(fvalues, set):
+                fvalues = set(fvalues)
 
             return _ensure_encoded(values), fvalues
         else: