python dataframe substract a column from multiple columns-CodePudding

I have a dataframe and I want to substract a column from multiple columns code:

df = pd.DataFrame('A':[10,20,30],'B':[100,200,300],'C':[15,10,50])

# Create a new A and B columns by sub-stracting C from A and B
df[['newA','newB']] = df[['A','B']]-df['C']

Present output:

    raise ValueError("cannot reindex from a duplicate axis")

ValueError: cannot reindex from a duplicate axis

CodePudding user response：

You can check sub

df[['newA', 'newB']] = df[['A', 'B']].sub(df['C'],axis=0)
df
Out[114]: 
    A    B   C  newA  newB
0  10  100  15    -5    85
1  20  200  10    10   190
2  30  300  50   -20   250

CodePudding user response：

Another option along with the above answer, you can convert column 'C' to a numpy array by doing df[['C']].values. Hence the new code would be:

df[['newA','newB']] = df[['A','B']]-df[['C']].values

CodePudding user response：

Try using Pandas .apply() method. You can pass columns and apply a given function to them, in this case subtracting one of your existing columns. The below should work. Documentation here.

df[['newA','newB']] = df[['A','B']].apply(lambda x: x - df['C'])

CodePudding user response：

You can try to convert df['C'].values to the same shape with df[['A','B']].

df[['newA','newB']] = df[['A','B']] - df['C'].values[:, None]

print(df)

    A    B   C  newA  newB
0  10  100  15    -5    85
1  20  200  10    10   190
2  30  300  50   -20   250