Merge pull request #634 from AK-ayush/num_cols

[REVIEW] Enable DataFrame creation from_pandas with numeric column names.

CHANGELOG.md

@@ -49,6 +49,7 @@
 - PR #612 Prevent an exception from occuring with true division on integer series.
 - PR #630 Fix deprecation warning for `pd.core.common.is_categorical_dtype`
 - PR #622 Fix Series.append() behaviour when appending values with different numeric dtype
+- PR #634 Fix create `DataFrame.from_pandas()` with numeric column names
 
 
 # cuDF 0.4.0 (05 Dec 2018)

python/cudf/dataframe/dataframe.py

@@ -6,6 +6,7 @@
 import random
 from collections import OrderedDict
 import warnings
+import numbers
 
 import numpy as np
 import pandas as pd
@@ -183,7 +184,7 @@ def __getattr__(self, key):
 
     def __getitem__(self, arg):
         """
-        If *arg* is a ``str``, return the column Series.
+        If *arg* is a ``str`` or ``int`` type, return the column Series.
         If *arg* is a ``slice``, return a new DataFrame with all columns
         sliced to the specified range.
         If *arg* is an ``array`` containing column names, return a new
@@ -217,7 +218,7 @@ def __getitem__(self, arg):
         >>> df[[True, False, True, False]] # mask the entire dataframe,
         # returning the rows specified in the boolean mask
         """
-        if isinstance(arg, str) or isinstance(arg, int):
+        if isinstance(arg, str) or isinstance(arg, numbers.Integral):
             s = self._cols[arg]
             s.name = arg
             return s
@@ -238,7 +239,7 @@ def __getitem__(self, arg):
             return df
         else:
             msg = "__getitem__ on type {!r} is not supported"
-            raise TypeError(msg.format(arg))
+            raise TypeError(msg.format(type(arg)))
 
     def __setitem__(self, name, col):
         """Add/set column by *name*

python/cudf/formatting.py

@@ -3,6 +3,7 @@
 """
 Define how data are formatted
 """
+import numbers
 
 
 def format(index, cols, show_headers=True, more_cols=0, more_rows=0,
@@ -49,7 +50,7 @@ def format(index, cols, show_headers=True, more_cols=0, more_rows=0,
     # compute column widths
     widths = {}
     for k, vs in cols.items():
-        if isinstance(k, int):
+        if isinstance(k, numbers.Integral):
             widths[k] = min_width
         else:
             widths[k] = max(len(k), max(map(len, vs), default=0), min_width)

python/cudf/tests/test_dataframe.py

@@ -187,6 +187,13 @@ def test_dataframe_column_name_indexing():
         for idx in combinations(pdf.columns, i):
             assert(pdf[list(idx)].equals(df[list(idx)].to_pandas()))
 
+    # test for only numeric columns
+    df = pd.DataFrame()
+    for i in range(0, 10):
+        df[i] = range(nelem)
+    gdf = DataFrame.from_pandas(df)
+    assert_eq(gdf, df)
+
 
 def test_dataframe_drop_method():
     df = DataFrame()