My organisation is running v13.2 of EM so there is no HPBNET node in EM and it has to be run through a code node. I have set the tool type to model and run the following code as per 'What's New in SAS Enterprise Miner 13.1':
proc HPBNET data=&em_import_data;
target %em_target;
input %em_nominal_input / level=nominal;
input %em_interval_input / level=interval;
code file="&em_file_empublishscorecode";
run;
The input is from a data partition node then metadata node. It runs without error but I get a warning that one of my variables is constant (which I don't understand because it has different values).
It seems to not produce a scored version of my training and validation datasets because I get errors when I try to run a model comparison node - ERROR: Variable P_FLAG_RCVR1 not found. (where P_FLAG_RCVR1 is the predicted value of my target variable.)
Any suggestions greatly appreciated. Or if there is more information that would help please let me know.
I have received an answer to this from tech support. Turns out the following code also needs to be added after the proc HPBNET statements: (This creates scored versions of datasets for comparison node.)
data &EM_EXPORT_TRAIN;
set &EM_IMPORT_DATA;
%inc "&EM_FILE_EMPUBLISHSCORECODE";
run;
data &EM_EXPORT_VALIDATE;
set &EM_IMPORT_VALIDATE;
%inc "&EM_FILE_EMPUBLISHSCORECODE";
run;
Would be much easier if we were running version 14.1 of EM 😉
I have received an answer to this from tech support. Turns out the following code also needs to be added after the proc HPBNET statements: (This creates scored versions of datasets for comparison node.)
data &EM_EXPORT_TRAIN;
set &EM_IMPORT_DATA;
%inc "&EM_FILE_EMPUBLISHSCORECODE";
run;
data &EM_EXPORT_VALIDATE;
set &EM_IMPORT_VALIDATE;
%inc "&EM_FILE_EMPUBLISHSCORECODE";
run;
Would be much easier if we were running version 14.1 of EM 😉
April 27 – 30 | Gaylord Texan | Grapevine, Texas
Walk in ready to learn. Walk out ready to deliver. This is the data and AI conference you can't afford to miss.
Register now and save with the early bird rate—just $795!
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.