A csv file with data on orders (for meals to be delivered) was provided, the documents comprises the folowing columns with information:
dateTime,restaurant,address,zippcodeFrom,zippcodeTo,dist,tm
With the format for dateTime like this: YYYY-MM-DD HH:MM:ss
I'd personally prefer to use MS Excel to apply the FFT (fast fourrier transform) to forecast based on time-series data. However, this is a python course, and the file is to large for MS Excel.
Getting the average number of orders per day of the week would be a start. But if I try the aggregation function, it sums all orders of all mondays alltogether.
How can i retrieve either the average number per weekday or the total number of mondays (and then divide the total number of orders on all mondays by the number of mondays? (Subsequently we have to do the same for the average total travel time (tm in the csv file) for delivery.
The challanage: multiple orders per day, result in multiple lines of data for each day. (The next thing is to get some kind of forecast hourly...)
What would be the best way to solve this?