CODEWITHSUNDEEP
pythonpandas
Laurent R
Laurent R

Reputation: 832

convert columns values into rows with pandas

I have a pd.DataFrame where I want to convert some columns into rows. I have the example below where I have 2 different sample with multiple target measurements. I want to break the targets ['t1', 't2', 't3'] and split them into new target row with the sample number. Is there a better way than a for-loop to convert a series of values (in columns) into rows ?

#The entry I have:
pd.DataFrame({'Sample':[0,1],
                  't1':[2,3],
                  't2':[4,5],
                  't3':[6,7]})

# the output I'm expecting:
pd.DataFrame({'Sample':[0,0,0,1,1,1],
              'targets':[2,4,6,3,5,7]})

I don't think that the pd.pivot_table() can do that for me. Does anyone have an idea ?

Upvotes: 1

Views: 74

Answers (1)

BENY
BENY

Reputation: 323226

You are looking for melt 

pd.DataFrame({'Sample':[0,1],
                  't1':[2,3],
                  't2':[4,5],
                  't3':[6,7]}).melt('Sample')
Out[74]: 
   Sample variable  value
0       0       t1      2
1       1       t1      3
2       0       t2      4
3       1       t2      5
4       0       t3      6
5       1       t3      7

Upvotes: 1

Related Questions