Regression on multiple items in different categories

[Majid2023]

Hi all,

I have a sample of 260 observations (52 weeks of sales for 5 items within two categories)

I would like to forecast weekly sales for each item for the next 4-6 week or forecast on weekly bases.

Is simple regression works here if yes how can I perform this task. 

Any help will be greatly appreciated.

[PaigeMiller]

I am extremely skeptical that a regression is a good tool to do this type of forecasting. I would think some time series model would work better. Nevertheless:

You need to add a record for weeks 53, 54, 55, 56, ... for each category and item to the data set, leaving sales as missing.

Then fit the regression using PROC REG with a BY statement (by category item;), and use the OUTPUT statement to ask for predicted values to be written to a SAS data set. The predicted values for weeks 53, 54, 55, 56, ... are the forecasts.

Here's a simple explanation, with code that would have to be modified for your actual application: https://blogs.sas.com/content/iml/2014/02/17/the-missing-value-trick-for-scoring-a-regression-model....

But really, you should be using a time series model.

[Majid2023]

I like your idea of time series, Can you help with time series code for this data set? 

[PaigeMiller]

To tell you the truth, I cannot help with time series models, other than the most simple ones. Probably someone else can help with that.

[Majid2023]

Thank you so much for sharing Regression code

[PaigeMiller]

You can also read the first 4 articles under Getting Started at this link https://documentation.sas.com/doc/en/pgmsascdc/9.4_3.4/etsug/etsug_arima_toc.htm which is an example of forecasting sales.

[Ksharp]

It is called panel data in time series analysis.
Better post it at Forecasting Forum:
https://communities.sas.com/t5/SAS-Forecasting-and-Econometrics/bd-p/forecasting_econometrics
Check PROC PANEL in SAS/ETS.

It is also called Longitudinal data in Clinic Data Analysis.
Check PROC MIXED or PROC GLIMMIX in SAS/STAT
https://blogs.sas.com/content/iml/2019/12/03/longitudinal-data-response-profile-model.html
https://blogs.sas.com/content/iml/2019/12/05/longitudinal-data-mixed-model.html

[Ksharp]

https://documentation.sas.com/doc/en/etsug/15.2/etsug_panel_examples05.htm