the following error occurs when i try to estimate the co-efficients for inputs_train and loan_data_targets_train using LogisticRegression:
Input contains NaN, infinity or a value too large for dtype('float64')
1 Answer
Hi Dan,
I ‘ve faced the same issue when I try regression part. The root cause of problem is inputs_train dataset have NaN valued column. you can check the null values in input_dataset columns inputs_train.isnull().sum()
I’ve followed the classroom codes and the main issue is caused by the creation of inputs_train_with_ref_cat variable. You should update the last column ‘mths_since_last_record:>=86’ to ‘mths_since_last_record:>86’ to run regression properly.
Hi Buğra, I’ve changed the last column ‘mths_since_last_record:>=86’ to ‘mths_since_last_record:>86’ to all the dataframes below and it works. Thanks. 🙂