BookmarkSubscribeRSS Feed
🔒 This topic is solved and locked. Need further help from the community? Please sign in and ask a new question.
CC_SAS
Obsidian | Level 7

Hi Community,

 

I searching for MASE formula used in SAS Forecast Studio. I googled but could find formula specific to SAS.

I know how it is computed in Python or R and more interested to know whether SAS uses the same formula or different.

The MASE formula which i used in python is from the below link:

https://otexts.com/fpp2/accuracy.html

 

FYI, we are using %fscreate() to run the forecasting jobs in SAS with model selection criteria as MASE.

 

Thank you

1 ACCEPTED SOLUTION

Accepted Solutions
Rick_SAS
SAS Super FREQ

Yes, that formula is included in the doc for Visual Forecasting.

 

Most of these formulas are standard and are the same across software products. What is the problem you are trying to solve? Are you getting different answers in SAS than you expected? If so, supply your data, the model you are using, and why you think the result is not correct.

View solution in original post

7 REPLIES 7
Rick_SAS
SAS Super FREQ

Google the documentation and you'll find it in the Appendix: Statistics of Fit:

Forecast Studio Doc

 

CC_SAS
Obsidian | Level 7

Hi Rick,

i have gone through the link which you shared. I could not find formula for MASE, just description is given.

Mean Absolute Scaled Error (MASE) - The mean of the absolute scaled errors.

I am looking for the formula, something similar to this

MASE=MAE / MAE(insample,naive)

 

Thanks

Rick_SAS
SAS Super FREQ

Yes, that formula is included in the doc for Visual Forecasting.

 

Most of these formulas are standard and are the same across software products. What is the problem you are trying to solve? Are you getting different answers in SAS than you expected? If so, supply your data, the model you are using, and why you think the result is not correct.

CC_SAS
Obsidian | Level 7

Hi Rick,

 

Thanks alot for the link.

I was looking for the below description:

For a non-seasonal series, the naïve forecast is generated by using the actual value from previous record. For a seasonal series, the naïve forecast is generated by using the actual value from previous season

 

Thanks again 🙂 

Rick_SAS
SAS Super FREQ

I don't think I have anything more to add. I merely Googled your keywords restricted to

   site:documentation.sas.com

 

Perhaps someone else can provide whatever further information you are looking for.

mitrov
SAS Employee

 

Like Rick said, the formula used by SAS is the same you linked to in Hyndman's book. They are standard definitions. Note that the link to the Visual Forecasting documentation he provided is for the weighted statistics across a group of series. This is the link to the documentation for a single series. It doesn't change much though, as they are obviously related. 

https://go.documentation.sas.com/?docsetId=vfug&docsetTarget=p0uawamu7dmtc2n1cllfwajyvlko.htm&docset...

 

My recollection is that FS only shows the weighted statistics over all series in the forecast summary, and maybe only WMAPE. I don't remember if it shows any other weighted statistics. 

 

If you are looking for the out-of-sample value of the statistics, you need to look at the FORECAST region. See also this:

https://go.documentation.sas.com/?docsetId=etsug&docsetTarget=etsug_esm_details10.htm&docsetVersion=...

 

SAS INNOVATE 2024

Innovate_SAS_Blue.png

Registration is open! SAS is returning to Vegas for an AI and analytics experience like no other! Whether you're an executive, manager, end user or SAS partner, SAS Innovate is designed for everyone on your team. Register for just $495 by 12/31/2023.

If you are interested in speaking, there is still time to submit a session idea. More details are posted on the website. 

Register now!

Multiple Linear Regression in SAS

Learn how to run multiple linear regression models with and without interactions, presented by SAS user Alex Chaplin.

Find more tutorials on the SAS Users YouTube channel.

Discussion stats
  • 7 replies
  • 1682 views
  • 1 like
  • 4 in conversation