Check with Timezone +2 Hrs GMT.
Thanks a lot @vbabu75, my data is now exactly fitting within the range BTW where did you find this information from, did I miss anything from the description.
For a single house/household_id I got the starting and ending timestamps. I converted them to find that they seemed to start the day before the first of the month at 10:00PM GMT and end on the last date of the month at 9:59PM GMT. So I guessed that the timestamps are in a timezone a couple of hours ahead of GMT.
I know the competition has ended but I still have a question. When I ingest data from input csv file, I noticed that I didn’t get data for lot of hours of date 20-09 and 21-09, for 21st there was only 3 hrs of data available.
But in alert_type_1 and _2 alerts for those missing time were there. How did other people solved this problem?
People how has got very good accuracy, I would be grateful if they can tell how did they counter this problem, It will definitely help in future competitions.
Thanks a lot