Hi there,
Does anyone know if the parameter defaults differ between running proc hpforest in say, EG and running through the node in Enterprise Miner (EM)?
I have copied as many parameters as I can find from my EM project and run proc hpforest in EG but am not getting the same model produced.
Grateful for any advice.
Thanks,
Martin
If you run the following in your Project Start Code in Enterprise Miner:
options mprint;
then run the HP Forest node, you can see in the SAS log in the Results what exactly runs, for example running the node with defaults I get this using the HMEQ data:
MPRINT(HPDM_HPFOREST): proc hpforest data=EMWS18.Ids2_DATA alpha=0.05 catbins=30 exhaustive=5000 maxdepth=50 maxtrees=100 mincatsize=5 missing=USEINSEARCH minUseInSearch=1 seed=12345 trainfraction = 0.6 ;
MPRINT(HPDM_HPFOREST): input
MPRINT(EM_INTERVAL_INPUT): CLAGE CLNO DEBTINC DELINQ DEROG LOAN MORTDUE NINQ VALUE YOJ
MPRINT(HPDM_HPFOREST): / level = interval;
MPRINT(HPDM_HPFOREST): input
MPRINT(EM_NOMINAL_INPUT): JOB REASON
MPRINT(HPDM_HPFOREST): / level = nominal;
MPRINT(HPDM_HPFOREST): * for the ordering of target levels;
MPRINT(HPDM_HPFOREST): * if not ASC, FMTASC, or FMTDESC then alway use DESC;
MPRINT(HPDM_HPFOREST): target BAD / level =
MPRINT(EM_TARGET_LEVEL): BINARY
MPRINT(HPDM_HPFOREST): order = DESCENDING;
MPRINT(HPDM_HPFOREST): ods output Baseline=EMWS18.HPDMForest3_BASELINE FitStatistics=EMWS18.HPDMForest3_ITERATION VariableImportance=EMWS18.HPDMForest3_VARIMPORT;
MPRINT(HPDM_HPFOREST): save file="E:\All_Users\wehass\EMProj\test\Workspaces\EMWS18\HPDMForest3\OUTMDLFILE.bin" ;
MPRINT(HPDM_HPFOREST): savestate file = "E:\All_Users\wehass\EMProj\test\Workspaces\EMWS18\HPDMForest3\score.sasast";
MPRINT(HPDM_HPFOREST): PERFORMANCE DETAILS ;
MPRINT(HPDM_HPFOREST): run;
If you run the following in your Project Start Code in Enterprise Miner:
options mprint;
then run the HP Forest node, you can see in the SAS log in the Results what exactly runs, for example running the node with defaults I get this using the HMEQ data:
MPRINT(HPDM_HPFOREST): proc hpforest data=EMWS18.Ids2_DATA alpha=0.05 catbins=30 exhaustive=5000 maxdepth=50 maxtrees=100 mincatsize=5 missing=USEINSEARCH minUseInSearch=1 seed=12345 trainfraction = 0.6 ;
MPRINT(HPDM_HPFOREST): input
MPRINT(EM_INTERVAL_INPUT): CLAGE CLNO DEBTINC DELINQ DEROG LOAN MORTDUE NINQ VALUE YOJ
MPRINT(HPDM_HPFOREST): / level = interval;
MPRINT(HPDM_HPFOREST): input
MPRINT(EM_NOMINAL_INPUT): JOB REASON
MPRINT(HPDM_HPFOREST): / level = nominal;
MPRINT(HPDM_HPFOREST): * for the ordering of target levels;
MPRINT(HPDM_HPFOREST): * if not ASC, FMTASC, or FMTDESC then alway use DESC;
MPRINT(HPDM_HPFOREST): target BAD / level =
MPRINT(EM_TARGET_LEVEL): BINARY
MPRINT(HPDM_HPFOREST): order = DESCENDING;
MPRINT(HPDM_HPFOREST): ods output Baseline=EMWS18.HPDMForest3_BASELINE FitStatistics=EMWS18.HPDMForest3_ITERATION VariableImportance=EMWS18.HPDMForest3_VARIMPORT;
MPRINT(HPDM_HPFOREST): save file="E:\All_Users\wehass\EMProj\test\Workspaces\EMWS18\HPDMForest3\OUTMDLFILE.bin" ;
MPRINT(HPDM_HPFOREST): savestate file = "E:\All_Users\wehass\EMProj\test\Workspaces\EMWS18\HPDMForest3\score.sasast";
MPRINT(HPDM_HPFOREST): PERFORMANCE DETAILS ;
MPRINT(HPDM_HPFOREST): run;
Good news: We've extended SAS Hackathon registration until Sept. 12, so you still have time to be part of our biggest event yet – our five-year anniversary!
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.