Fractionally_dff features and , min_max standardization

AI & Machine learning
1
yet here I would like to know if you reshape your data between (0,1) or any other appropriate values after fractionally differentiating your features ? I have tried but it is always that ml classifiers produces error mainly saying that input data might be too large or have infinite values , or it contains NANS , though I have cleaned all of those using the .sum , .count , isnull() functions , column by column including my target variable
 
Your input is highly appreciated 
best regards
ashraf
2

Hello Ashraf,



How about trying the following template code for looking for all rows in all columns that aren't trainable in ml models… i.e they are nan or large values…



df[~df.isin([np.nan, np.inf, -np.inf]).any(1)]



Then perhaps you can replace or get rid of these indices.