Add Series.name and Series.from_pandas

Author: mrocklin

pygdf/dataframe.py

@@ -423,6 +423,7 @@ def add_column(self, name, data, forceindex=False):
             raise NameError('duplicated column name {!r}'.format(name))
 
         series = self._prepare_series_for_add(data, forceindex=forceindex)
+        series.name = name
         self._cols[name] = series
 
     def drop_column(self, name):

pygdf/series.py

@@ -6,6 +6,7 @@
 from numbers import Number
 
 import numpy as np
+import pandas as pd
 
 from . import cudautils, formatting
 from .buffer import Buffer
@@ -59,9 +60,20 @@ def from_masked_array(cls, data, mask, null_count=None):
         return cls(data=col)
 
     def __init__(self, data, index=None):
+        name = None
+        if isinstance(data, pd.Series):
+            from .dataframe import DataFrame
+            name = data.name
+            data = data.to_frame()
+            data.columns = ['x']
+            data = DataFrame.from_pandas(data)
+            data = data['x']
+            data.name = name
         if isinstance(data, Series):
             index = data._index
+            name = data.name
             data = data._column
+
         if not isinstance(data, columnops.TypedColumnBase):
             data = columnops.as_column(data)
 
@@ -71,6 +83,11 @@ def __init__(self, data, index=None):
         assert isinstance(data, columnops.TypedColumnBase)
         self._column = data
         self._index = RangeIndex(len(data)) if index is None else index
+        self.name = name
+
+    @classmethod
+    def from_pandas(cls, s):
+        return cls(s)
 
     def serialize(self, serialize):
         header = {}
@@ -448,7 +465,9 @@ def to_gpu_array(self, fillna=None):
     def to_pandas(self, index=True):
         if index is True:
             index = self.index.to_pandas()
-        return self._column.to_pandas(index=index)
+        s = self._column.to_pandas(index=index)
+        s.name = self.name
+        return s
 
     @property
     def data(self):

pygdf/tests/test_dataframe.py

@@ -670,3 +670,23 @@ def do_slice(x):
     got = do_slice(gdf).to_pandas()
 
     pd.testing.assert_frame_equal(expect, got)
+
+
+def test_from_pandas():
+    df = pd.DataFrame({'x': [1, 2, 3]}, index=[4., 5., 6.])
+    gdf = gd.DataFrame.from_pandas(df)
+    assert isinstance(gdf, gd.DataFrame)
+
+    pd.testing.assert_frame_equal(df, gdf.to_pandas())
+
+    s = df.x
+    gs = gd.Series.from_pandas(s)
+    assert isinstance(gs, gd.Series)
+
+    pd.testing.assert_series_equal(s, gs.to_pandas())
+
+
+def test_series_name():
+    df = pd.DataFrame({'x': [1, 2, 3]}, index=[4., 5., 6.])
+    gdf = gd.DataFrame.from_pandas(df)
+    assert gdf['x'].name == 'x'