s_khan92 :
I have a large dataset where i have about more than 350 columns and i want to convert that 350 columns into rows and rows values as columns. The users provide the ranking "1-10" to brands and i want to show ranking(1-10) as coluumns. Here is how my df looks:
id Gender Age BMW Audi VW Mercedes
1 M 20 3 1 4 2
2 F 50 1 2 3 4
3 F 25 2 1 4 3
And i would like to change it in this way as desired column.
id Gender Age Rank1 Rank2 Rank3 Rank4
1 M 20 Audi Mercedes BMW VW
2 F 50 BMW Audi VW Mercedes
3 F 25 Audi BMW Mercedes VW
In this way i can get rid of so many columns... Any solution?
anky_91 :
you can melt+pivot:
idx = ['id','Gender','Age']
m = df.melt(idx).pivot_table(index=idx,columns='value',values='variable',aggfunc='first')
out = m.add_prefix('Rank').reset_index(idx).rename_axis(index=None,columns=None)
print(out)
id Gender Age Rank1 Rank2 Rank3 Rank4
0 1 M 20 Audi Mercedes BMW VW
1 2 F 50 BMW Audi VW Mercedes
2 3 F 25 Audi BMW Mercedes VW
Guess you like
Origin http://10.200.1.11:23101/article/api/json?id=390920&siteId=1