['AfterStateHoliday']: Input contains NaN, infinity or a value too large for dtype('float32')

Part 1 (2018)
22

The notebook update didn’t fix this error and I used the same two lines of fillna with inplace=True to fix the problem. Private score 0.26106 Public score
0.10199

2 Likes
23

Yes can confirm that the latest code in GIT after the fix for get_elapsed runs fine in terms of the error as long as you run df separately for train and test, thereby making sure that joined and joined_test are created correctly. Thanks

2 Likes