CODEWITHSUNDEEP
pythonpandas
Si_CPyR
Si_CPyR

Reputation: 161

How to create multiple subsets of DataFrame with loop

Base DataFrame

df = pd.DataFrame(np.random.randint(0,10,size=(100,6)),columns = 
['red','blue','yellow','green','purple','total'])

How do I code this with a loop?

df_x = df[df.columns[[y,5]]]

I'd like to apply this logic to a dataframe with many more columns, but for simplicity sake, boiled down the question into this.

Below is desired output, 5 new dataframes created (Assuming I don't have to type all the 5 lines)

df_red = df[df.columns[[0,5]]]
df_blue = df[df.columns[[1,5]]]
df_yellow = df[df.columns[[2,5]]]
df_green = df[df.columns[[3,5]]]
df_purple = df[df.columns[[4,5]]]

Upvotes: 0

Views: 502

Answers (1)

BENY
BENY

Reputation: 323276

You can do with groupby axis =1 and pd.concat

variables = locals()
for x , y  in df.iloc[:,:-1].groupby(level=0,axis=1):
    variables["df_{0}".format(x)] =pd.concat([y,df.iloc[:,[-1]]],axis=1)


df_red.head()

Out[566]: 
   red  total
0    4      7
1    7      9
2    6      7
3    4      2
4    5      8

Upvotes: 2

Related Questions