DataFrame.fillna corrupts columns with duplicated names

[hantusk]

# Pandas version 0.17.1
import pandas as pd
df = pd.DataFrame({'Same': 1.0, ' Same': pd.np.nan, '  Same': pd.np.nan}, index=[0,1,2])
df.columns = [c.strip() for c in df.columns]
df.iloc[:, 2] # Returns all 1.0

df.iloc[:, 0] = df.iloc[:, 0].fillna(df.iloc[:, 1])

df.iloc[:, 2] # Column 2 is corrupted and returns all NaN

[jreback]

hmm, that appears to be the case. pull-requests to investigate are welcome. Duplicate column support should work, though not as complete as more standard unique support.

[gfyoung]

@jreback : Where would the potentially offending code be located?

[jreback]

core/generic