CODEWITHSUNDEEP
pythonpandasnumpy
user14815110
user14815110

Reputation: 133

How to use column name as a variable?

I have a dataframe as below.

d={
    'id' :[1,2,3,4,5],
    'X1':[25,12,3,7,1],
    'X2':[3,45,3,1,6],
    'X3':[6,23,9,3,2]
}

df=pd.DataFrame(d, columns=['id','X1','X2','X3'])

looks like this:

enter image description here

I want to add a need column as X4. But I want to use column name and its value as below. How to use column name as a variable?

Expected X4 value for first raw:

X1:25,X2:3,X3:6

Upvotes: 1

Views: 583

Answers (4)

ggaurav
ggaurav

Reputation: 1804

A different general approach:

def get_other_cols_vals(df1):
    vals = []
    for i_enm, col in enumerate(df1.columns):
        vals.append(f"{col}:{df1.iloc[0, i_enm]}")
    return ",".join(vals)

df['X4'] = df.groupby(df.index)[df.columns.drop('id').values].apply(get_other_cols_vals)

enter image description here

Upvotes: 1

Jarvis
Jarvis

Reputation: 8564

This should do:

df['X4'] = 'X1:' + df['X1'].astype(str) + ',X2:' + df['X2'].astype(str) + ',X3:' + df['X3'].astype(str)

gives

   id  X1  X2  X3                 X4
0   1  25   3   6    X1:25,X2:3,X3:6
1   2  12  45  23  X1:12,X2:45,X3:23
2   3   3   3   9     X1:3,X2:3,X3:9
3   4   7   1   3     X1:7,X2:1,X3:3
4   5   1   6   2     X1:1,X2:6,X3:2

A more general and dynamic solution where you can just specify a list of columns as an argument to df[...].apply():

df['X4'] = df[['X1', 'X2', 'X3']].apply(
               lambda x: ','.join(df.columns[x.name+1] + x.astype(str)), axis=1
           )

Upvotes: 2

Chris
Chris

Reputation: 16147

d={
    'id' :[1,2,3,4,5],
    'X1':[25,12,3,7,1],
    'X2':[3,45,3,1,6],
    'X3':[6,23,9,3,2]
}

df=pd.DataFrame(d, columns=['id','X1','X2','X3'])

cols = [col for col in df.columns if 'X' in col]
df['X4'] = df[cols].apply(lambda x: ';'.join([str(x)+":"+str(y) for x,y in zip(cols,x.values)]), axis=1)

Output

    id  X1  X2  X3  X4
0   1   25  3   6   X1:25;X2:3;X3:6
1   2   12  45  23  X1:12;X2:45;X3:23
2   3   3   3   9   X1:3;X2:3;X3:9
3   4   7   1   3   X1:7;X2:1;X3:3
4   5   1   6   2   X1:1;X2:6;X3:2

Upvotes: 2

Rob Raymond
Rob Raymond

Reputation: 31146

Create a new column X4 that is the composition you want.

d={
    'id' :[1,2,3,4,5],
    'X1':[25,12,3,7,1],
    'X2':[3,45,3,1,6],
    'X3':[6,23,9,3,2]
}

df=pd.DataFrame(d, columns=['id','X1','X2','X3'])
df = df.assign(X4=lambda dfa: "X1:" + dfa["X1"].astype("string")
         + ",X2:" + dfa['X2'].astype("string")
         + ",X3:" + dfa['X3'].astype("string")
         )

output

 id  X1  X2  X3                 X4
  1  25   3   6    X1:25,X2:3,X3:6
  2  12  45  23  X1:12,X2:45,X3:23
  3   3   3   9     X1:3,X2:3,X3:9
  4   7   1   3     X1:7,X2:1,X3:3
  5   1   6   2     X1:1,X2:6,X3:2

Upvotes: 2

Related Questions