如何在Pandas中使用group-by？

Question

My data-frame look like this(two columns col1,col2)

1 100
2 150
3 170
1 200

I want to do group-by col1,

pd.DataFrame(combined.groupby('col1').aggregate(np.mean)['col2'])

This is returning a data-frame with only one key col2, I, actually, want the output to be like this( dataframe with two columns)

col1,mean(col2), 

could somebody point out what do I have to achive this?

Answer 1

You can use groupby with aggregating mean and reset_index :

print df.groupby('col1')['col2'].mean().reset_index()
   col1  col2
0     1   150
1     2   150
2     3   170

Solution with groupby with paameter as_index=False as mentioned John Galt :

print df.groupby('col1', as_index=False)['col2'].mean()
   col1  col2
0     1   150
1     2   150
2     3   170

Solution with aggregate :

print df.groupby('col1', as_index=False).aggregate({'col2':'mean'})
   col1  col2
0     1   150
1     2   150
2     3   170

Aggregation in docs