BookmarkSubscribeRSS Feed
smita249
Calcite | Level 5

Hi,

 

 

I am using the cluster node in Enterprise Miner 14.1. There are 6 segments produced but I see only 3 segment variables instead of 6 in the segment plot. What changes do I have to make to the default options to see all. 

 

Please see a snapshot of the segment pie chart and the plot in the atachment. 

 

emchart.png

 

Thanks!

 

6 REPLIES 6
WendyCzika
SAS Employee

We aren't sure why you aren't seeing all segments in the top plot.  You can look at the data that the plot is using by clicking on the Data icon when that plot is selected and see if that gives any clues about why it's only showing 3 segments - let us know!

smita249
Calcite | Level 5

Hi,

 

Thanks for your suggestion. Here is the screenshot of the data options. Nothing seems relevant pertaining to the graph.

 

Capture.JPG

 

 

Here is the screenshot of the graph options. The default option here for the maximum bars is 3 - grayed out.

 

Capture1.JPG

 

 

Also see the axes option in graph properties .. it is grayed out - automatic from 0 to 2.

 

Capture2.JPG

 

Thanks!

WendyCzika
SAS Employee

This is the icon I mean (highlighted below) to see the data used in the plot:

 

Cluster.PNG

smita249
Calcite | Level 5

Hi,

 

I looked at it - it shows only 3 segment variables, not 6.

 

Thanks! 

WendyCzika
SAS Employee

I can't figure out why that is happening - we can't reproduce that with other data.  As a workaround, you can try using a Segment Profile node (from the Assess tab) to help you view relationships between your segments and variables.  Hopefully you will see all segments there (be sure to set Profile All property to Yes).

DougWielenga
SAS Employee

It would be very helpful if you could try a couple things that would let us know if this issue is specific to the installation (e.g. something not configured correctly) or if there is something specific to the data (e.g. certain variables have too much missing information). It is also possible that the node and/or diagram might have gotten corrupted.  Have you tried adding a new Cluster node to the flow so to see if it generates the same results?  You do not need to delete the old Cluster node, just add a new node following the other Cluster node's predecessor, run it, and view the results.  

 

If you obtain the same results from the new node/flow, please do the following:

1 - Create a new SAS Enterprise Miner project

2 - Click on Help --> Generate Sample Data Sources... 

3 - Make sure the Home Equity data set is selected and press OK

4 - Add a new diagram to the project

5 - Add the new Home Equity data source to the diagram

6 - Connect a default Cluster node following the Home Equity data source

7 - Run the flow and view the results

 

Does it show the correct number of segments?  If so, we will need to look at your data source to better understand what might be happening.  If not, there is likely an issue with your installation (e.g. perhaps it is using the wrong version of Java, etc...).   Regardless of what happens, please create a model package from the Cluster node following the Home Equity data set and another model package from the new Cluster node you added following your original data set.  Please send the model package using the sample data to me.   It would also be helpful to review the model package from your original data to see what is happening.  


To create a model package, follow the steps in SAS Usage Note 46764 available at 

 

    http://support.sas.com/kb/46/764.html

 

I look forward to hearing from you!


Cordially,

Doug

sas-innovate-2024.png

Join us for SAS Innovate April 16-19 at the Aria in Las Vegas. Bring the team and save big with our group pricing for a limited time only.

Pre-conference courses and tutorials are filling up fast and are always a sellout. Register today to reserve your seat.

 

Register now!

How to choose a machine learning algorithm

Use this tutorial as a handy guide to weigh the pros and cons of these commonly used machine learning algorithms.

Find more tutorials on the SAS Users YouTube channel.

Discussion stats
  • 6 replies
  • 1400 views
  • 0 likes
  • 3 in conversation