按列对Pandas DataFrame进行排序

Question

How do I sort a pandas data-frame by column? I read here http://pandas.pydata.org/...sort_values about sort_values. When I run

df = pd.DataFrame([(2,'a'),(1,'b')],columns = ['num','let'])
df
df.sort_values(by='num', axis=1, ascending=True, inplace=True)

I obtain error KeyError: 'num'

Answer 1

You should remove axis=1 , this tries to look for index labels that match 'num' which don't exist:

In[33]:
df.sort_values(by='num')

Out[33]: 
   num let
1    1   b
0    2   a

The default is axis=0 which looks for columns that match 'num'