topic Re: Anyone still using the Treedisc macro? in SAS/IML Software and Matrix Computations

Anyone still using the Treedisc macro?

BobtheModelBuilder — Thu, 17 Apr 2014 16:07:50 GMT

Hello,

Is anyone still using the treedisc macro SAS released so many years ago? I used to have a nice little program that used it and the xmacro.sas program to run a basic CHAID analysis. It wasn't slick but did the job when I needed it. However, sometime after I got upgraded to version 9.3 it doesn't seem to work anymore. The program errors out in the IML section and I don't know IML enough to debug. Does anyone still running Treedisc.sas have any problems with running it in V9.3?

Thanks!

Re: Anyone still using the Treedisc macro?

art297 — Thu, 17 Apr 2014 17:28:15 GMT

I used it as recently as last June and, as best I recall, we were running on SAS 9.3 at the time.

Re: Anyone still using the Treedisc macro?

Hutch_sas — Thu, 17 Apr 2014 17:56:28 GMT

Maybe if you posted the code and the error log we might be able to figure out what is causing the error?

Re: Anyone still using the Treedisc macro?

BobtheModelBuilder — Thu, 17 Apr 2014 21:19:17 GMT

I will have to run it again with the macro tracing on and save the log to see where the error occurs.

Re: Anyone still using the Treedisc macro?

SASDescUser — Fri, 20 Jun 2014 05:57:56 GMT

Hi,

I am trying to run the %TREEDISC macro of SAS on SAS Enterprise Guide 4.3.

I have obtained the two codes from the following links.

Treedisc.sas: http://www.public.iastate.edu/~kkoehler/stat557/sas/treedisc.sas/
Xmacro.sas: http://schick.tripod.com/xmacro.sas/

I have changed the following statement to refer to the xmacro.sas-the file containing XMACRO on my system.

*%inc '/sas/m611/stat/sampsrc/xmacro.sas';

I am trying to execute the example given in the beginning of treedisc.sas code to run the %treedisc macro.

I get the following errors after which the treedisc macro terminates.

17 *** Compute a tree for predicting SPECIES from the petal and

18 sepal lengths and widths, which are treated as ordinal

19 predictors;

20 %treedisc( data=iris, depvar=species, ordinal=petal: sepal:,

21 outtree=trd, options=noformat, trace=2);

ERROR: Procedure IML not found.

ERROR: Unrecognized form of the RUN statement. Use either RUN; or RUN CANCEL;.

ERROR: File WORK._DESC_.DATA does not exist.

WARNING: The data set WORK._DESC_ may be incomplete. When this step was stopped there were 0 observations

and 2 variables.

ERROR: Unknown error constructing OUTTREE= data set while running the TREEDISC macro.

NOTE: The TREEDISC macro terminated due to error(s).

Can anyone please guide me to fix these issues?