Tuesday, August 30, 2022

Outlier Detecting in Pandas

 Q1 = df_boston['CRIM'].quantile(0.25)

Q3 = df_boston['CRIM'].quantile(0.75)
IQR = Q3 - Q1

Outlier_min = Q1 - 1.5 * IQR 
Outlier_max = Q3 + 1.5 * IQR
print(IQR, Outlier_min, Outlier_max)


###outlier dealing
df_boston['CRIM'] = np.where(df_boston['CRIM']>= Outlier_max, Outlier_max,df_boston['CRIM'])
df_boston['CRIM'] = np.where(df_boston['CRIM']<= Outlier_min, Outlier_min,df_boston['CRIM'])

No comments:

Post a Comment