Please my graph for Proc Cluster is not showing

[tojojosh]

Please assist, this is the code I used for my analysis however I cant see the display of the cluster graph

FILENAME REFFILE '/folders/myfolders/lesson3/Electronic.xlsx';

PROC IMPORT DATAFILE=REFFILE
    DBMS=XLSX
    OUT=WORK.Electronic;
    GETNAMES=YES;
RUN;

ods graphics on
Proc Cluster Data=Electronic Print=7 Simple Method=Centroid rmsstd rsquare outtree=Electronic_Dataset plots(maxpoints=5000);
 Id Order_ID;
 Var Sales Profit;
 Run;
 ods graphics off ;

The warning its giving me is as follows

WARNING: Ties for minimum distance between clusters have been detected at 1272 level(s) in the cluster history.

WARNING: DENDROGRAM statement will not be drawn because the DISCRETEMAX threshold has been reached. You can set DISCRETEMAX=1,400

on the ODS GRAPHICS statement to draw the plot.

WARNING: DENDROGRAM statement will not be drawn because the DISCRETEMAX threshold has been reached. You can set DISCRETEMAX=1,400

on the ODS GRAPHICS statement to draw the plot.

WARNING: DENDROGRAM statement will not be drawn because the DISCRETEMAX threshold has been reached. You can set DISCRETEMAX=1,400

on the ODS GRAPHICS statement to draw the plot.

[Rick_SAS · Accepted solution]

The WARNING message tells you the solution. Modify the ODS GRAPHICS statement to read:

ods graphics on / DISCRETEMAX=1400;

then run the procedure again.

[tojojosh]

Thanks Rick, I was omitting the '/' sign earlier.
Thank you

[tojojosh]

Thanks Rick, I can see the graph now but I am still getting the ff warning: Ties for minimum distance between clusters have been detected at 1272 level(s) in the cluster history.

Pls is this an issue ?

[Rick_SAS]

That is a data issue. It means that some observations are the same distance from two or more cluster centers. This situation can occur when you have discrete values in your data. (For example, if salaries and profits are rounded to the nearest $1000.) It indicates that an observation is equally likely to belong to more than one cluster.  

There is nothing you can do about this warning unless you change your data to use more precision (for example, measure salaries and profits to the nearest cent.)

[tojojosh]

Thanks a bunch