Melt with Pandas



In [3]:

    
import pandas as pd



In [5]:

    
df = pd.DataFrame([
    {'first': 'john', 'last': 'smith', 'height': '180', 'weight': '80'},
    {'first': 'jane', 'last': 'doe', 'height': '160', 'weight': '50'}
])
df.head()









    Out[5]:






  
    
      
      first
      height
      last
      weight
    
  
  
    
      0
      john
      180
      smith
      80
    
    
      1
      jane
      160
      doe
      50



In [6]:

    
pd.melt(df, id_vars=['first', 'last'])









    Out[6]:






  
    
      
      first
      last
      variable
      value
    
  
  
    
      0
      john
      smith
      height
      180
    
    
      1
      jane
      doe
      height
      160
    
    
      2
      john
      smith
      weight
      80
    
    
      3
      jane
      doe
      weight
      50



In [7]:

    
pd.melt(df, id_vars=['first'])









    Out[7]:






  
    
      
      first
      variable
      value
    
  
  
    
      0
      john
      height
      180
    
    
      1
      jane
      height
      160
    
    
      2
      john
      last
      smith
    
    
      3
      jane
      last
      doe
    
    
      4
      john
      weight
      80
    
    
      5
      jane
      weight
      50

Here's a rel-life example.

	first	last	variable	value
0	john	smith	height	180
1	jane	doe	height	160
2	john	smith	weight	80
3	jane	doe	weight	50