I have a dataframe with 2 columns X and Y. I want to set up a new series which equals X when X=Y, and x/(Y-X) when X!=Y.
If try to do this:
df['temp'] = 0
df.loc[df.X == df.Y, 'temp'] = df.X
df.loc[df.X != df.Y, 'temp'] = df.X /(df.Y-df.X)
Z= df.temp
df = df.drop('temp', axis=1)
I get
ZeroDivisionError: float division by zero
Is there a better way to do this than running loops?
Data
df=pd.DataFrame({'X':[1,2,3,4,5], 'Y':[2,2,6,4,5]})
df
Use
Import numpy as np
df=df.assign(j=(np.where(df.X==df.Y, df.X, df.X.div(df.Y-df.X))))
df
Sol
X Y j
0 1 2 1.0
1 2 2 2.0
2 3 6 1.0
3 4 4 4.0
4 5 5 5.0
The condition should apply to both side
df['temp'] = 0
df.loc[df.X == df.Y, 'temp'] = df.X
df.loc[df.X != df.Y, 'temp'] = df.loc[df.X != df.Y,'X'] /(df.loc[df.X != df.Y,'Y']-df.loc[df.X != df.Y,'X'])
Z= df.temp
df = df.drop('temp', axis=1)
Try:
divisor = np.where(df.Y==df.X, 1, df.Y-df.X)
Z = df.X / divisor
