I noticed that the 'Sample_Submission.csv' file is missing many hours in the 24-hour sequence over the 56-day prediction period (from July 13th to September 6th). If the call center is open 24 hours every day, the prediction would be for 1344 hours (56*24); however, I note that the prediction hours specified in the 'Sample_Submission.csv' only covers 1184 hours. For example, I find that there is a gap between ' 2016071516' and '2016071809'. Is this intentional? If so, I hadn't accounted for this initially and will need to modify the test submission output.
Sample submission is not the test dataset.It is actually included to show you how your submission file should look
Thank you. Just to clarify, the omission of hours from the sample submission csv file is intended? So, our submission would also skip these hours as well? Thanks!
No.All the hours witghin the specified range Must be included.You can use date_ranage to generate the the dates and specifiy your frequency has hours.For example date_rng = pd.date_range(start='13/07/2016', end='16/09/2016', freq='H') generates all hours between 13th Jult and 16th oct.Hope this helps!