Alteryx Designer Desktop Discussions

Find answers, ask questions, and share expertise about Alteryx Designer Desktop and Intelligence Suite.
SOLVED

How to do a forcast for multiple companies

IvanMtz
6 - Meteoroid

Hello Community 

 

I have a question is just that I need to do a forecast for each company that we do business I have a file that is as an example like this:

 

Company # |  Company Name |  Amount Paid | Payment Date |

11111111         Company 1                  10                10/08/2021

11111112         Company 2                  11                10/08/2021

11111111         Company 1                  15                10/08/2021

11111113         Company 3                  30                10/08/2021

11111111         Company 1                  10                10/08/2021

11111111         Company 1                  10                10/08/2021

11111112         Company 2                  11                10/08/2021

11111111         Company 1                  15                10/08/2021

11111113         Company 3                  30                10/08/2021

11111111         Company 1                  10                10/08/2021

 

With what tool I can do a forecast without filtering per company and do a forecast for each company.

 

Thank you in advance 

 

Have a nice weekend!

 

 

3 REPLIES 3
danilang
19 - Altair
19 - Altair

Hi @IvanMtz 

 

None of the Time Series tools allow you to group by.  You'll have to use a batch macro.  Here's the basic structure 

 

danilang_0-1633791386827.png

 

The control parameter modifies the Filter to select records for the current company and also modifies the Model Name attribute in the ETS tool so you'll be able to associate the values with the companies.

 

Note that the attached doesn't run as is, since your input data only has values for one date.

 

Dan 

 

NeilR
Alteryx Alumni (Retired)

Also check out the Time Series Factory tools. 

LiuZhang
9 - Comet

Given you have only one additional column on company, have you done a line/box plot just to see if there are different in the first place?

 

If they are different you could try to use covariate in arima model to see if it improves your prediction. 

Labels