I am having a big database with entries of a player count and a datetime (see [login to view URL]) - based on these values, I am drawing a chart ([login to view URL]).
I now need a "future drawing" for the chart, which is predicting the potential player count based on the next hours.
The prediction should be made by considering the values of the same time frame of the last 7 days.
As per hour I have more values, example:
7:03pm = 100 players
7:07pm = 89 players
7:19pm = 99 players
7:47pm = 105 players
you should find a proper way to give them a weighed average for the time between 7pm and 8pm and then use these values of each hour of the last 7 days, to give an estimation of the future timeperiod.
Example1:
Day 1 - 7:00pm = 100 players
Day 2 - 7:00pm = 80 players
Day 3 - 7:00pm = 95 players
Day 4 - 7:00pm = 85 players
Day 5 - 7:00pm = 72 players
Day 6 - 7:00pm = 90 players
Day 7 - 7:00pm = 80 players
Prediction of Day 8 - 7:00 pm should be something like 86, as it would be the average value of the last days at the same time.
The prediction should be made always for the next 6 hours. You dont need to export a chart, just the values and time is fine - Example:
Prediction 8:00pm = 30
Prediction 9:00pm = 28
Prediction 10:00pm = 23
Prediction 11:00pm = 20
Prediction 12:00pm = 15
Prediction 1:00am = 9
If you have any questions, please let me know. I can provide the database entries for you to test.