i used robustscaling and labelencoding for numerical and categorical column respectively at the time of training.
num_features = x_df[numeric_type]
scaler = RobustScaler()
x_df[numeric_type] = scaler.fit_transform(num_features.values)
label_encoder = LabelEncoder()
x_df[categorical_type[i]] = label_encoder.fit_transform(x_df[categorical_type[i]])
and it works very fine.
and i trained my model also but when i try to give new values for prediction so i have to give value in the encoded aur scaled form
eg: at training : apple -> 1
at prediction: if i want to insert apple as a value so i want to type apple but this says you have to give 1 so any solution to encode with same technique as used in training with the sequence of variable also. means variable will user defined so i dont have idea of sequence of variable what user decided for training and prediction. thanks in advance
