If you get the below error while predicting missing values, You should use SimpleImputer instead of Imputer.
DeprecationWarning: Class Imputer is deprecated; Imputer was deprecated in version 0.20 and will be removed in 0.22. Import impute.SimpleImputer from sklearn instead.
A sample code that show how to use SimpleImputer is given below.
# Importing the libraries import numpy as np import matplotlib.pyplot as plt import pandas as pd # Importing the dataset dataset = pd.read_csv('my_data.csv') # Get all independent data marix, all columns except the last one Independent = dataset.iloc[:, :-1].values # Our original matrix has 5 columns, so the last column is #dependent and index of that one is 4 Dependent = dataset.iloc[:, 4].values # Fill the missing data using SimpleImputer # Earlier Imputer was uisng for this, but now we have to use # SimpleImputer # from sklearn.preprocessing import Imputer from sklearn.impute import SimpleImputer # Old code that is deprecated now. #imputer = Imputer(missing_values = 'NaN', strategy = 'mean', axis = 0) # new code that will work imputer = SimpleImputer(missing_values=np.nan,strategy='mean') imputer = imputer.fit(Independent[:, 1:4]) Independent[:, 1:4] = imputer.transform(Independent[:, 1:4]) print(Independent)