Want to get involved? We're always looking for ideas and content for Weekly Challenges.
SUBMIT YOUR IDEAI have to say this one wound me up! I wasted so much time due to diligently converting the YEAR and MONTH columns into a date, only to realise the time series tools require the data in that 2 column format... Why wouldn't the time series tools accept proper date formats?!
Also to then have to tell the tool on what year and month the incoming data starts manually seems absurd. Should have been built in to the macro.
My answers vary a little depending on whether you account for the first record by doing the running sum before or after the time series forecast.
Forgot to union in the original data to do the running sum after the prediction, I guess this is the better approach, seemed to be what the solution was looking for anyway.
Thank goodness you can look at other people's solutions eventually, it took me about an hour...!