BookmarkSubscribeRSS Feed
☑ This topic is solved. Need further help from the community? Please sign in and ask a new question.
ogrigny
Fluorite | Level 6

Hi,

I have developed a segmentation with eight segments that I am satisfied with and I want to follow the evolution of my customers over time through these 8 segments. How can I make sure that when I update my segmentation, I keep the same segments as before (for example, that segment 8 still has the same characteristics as the previous segment 8 and that only my customers have moved from one segment to the other)? I use the fastclust procedure, does using the outseed options to retrieve the initial seeds and then use the seed option when updating keep the same segments? Thanks in advance for your help!

1 ACCEPTED SOLUTION

Accepted Solutions
KevinScott
SAS Employee

Yes, this can be done with PROC FASTCLUS:

 

There is an example In technical note 22544 here: 22544 - Assigning new observations to clusters defined using previous data (sas.com)

I will reiterate the relevant information below:

 

Assigning new observations to clusters with PROC FASTCLUS using a cluster solution from PROC FASTCLUS

Suppose you have an eight-cluster solution for your original data and saved the cluster solution statistics in the OUTSTAT= data set FastclusStats.

 

This dataset will be used as the INSTAT= data set to assign the new observations you receive over time to the existing eight clusters. Your final cluster assignments will be saved in the NewClusters data set.

 

proc fastclus data=NewData instat=FastclusStats out=NewClusters;
  var <yourvars>;
run;

 

Any transformation to your variables before the initial execution of FASTCLUS to find the eight clusters will need to be done to the same variables in the new data set that you use to assign/score observations into the existing eight clusters.

 

After the reassignment of customers, you can do post-processing to see which customers have changed clusters and aggregate the migration/transition measures

View solution in original post

1 REPLY 1
KevinScott
SAS Employee

Yes, this can be done with PROC FASTCLUS:

 

There is an example In technical note 22544 here: 22544 - Assigning new observations to clusters defined using previous data (sas.com)

I will reiterate the relevant information below:

 

Assigning new observations to clusters with PROC FASTCLUS using a cluster solution from PROC FASTCLUS

Suppose you have an eight-cluster solution for your original data and saved the cluster solution statistics in the OUTSTAT= data set FastclusStats.

 

This dataset will be used as the INSTAT= data set to assign the new observations you receive over time to the existing eight clusters. Your final cluster assignments will be saved in the NewClusters data set.

 

proc fastclus data=NewData instat=FastclusStats out=NewClusters;
  var <yourvars>;
run;

 

Any transformation to your variables before the initial execution of FASTCLUS to find the eight clusters will need to be done to the same variables in the new data set that you use to assign/score observations into the existing eight clusters.

 

After the reassignment of customers, you can do post-processing to see which customers have changed clusters and aggregate the migration/transition measures

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!

What is Bayesian Analysis?

Learn the difference between classical and Bayesian statistical approaches and see a few PROC examples to perform Bayesian analysis in this video.

Find more tutorials on the SAS Users YouTube channel.

Get the $99 certification deal.jpg

 

 

Back in the Classroom!

Select SAS Training centers are offering in-person courses. View upcoming courses for:

View all other training opportunities.

Discussion stats
  • 1 reply
  • 184 views
  • 1 like
  • 2 in conversation