Before you analyze the data using the DMINE procedure, you must create
the DMDB encoded data set and catalog. For more information about how to do
this, see "Example 1: Getting Started with the DMDB Procedure"
in the DMDB procedure documentation.
proc dmdb batch data=sampsio.dmexa1 out=dmbexa1 dmdbcat=catexa1;
id acctnum;
var amount income homeval frequent recency age
domestic apparel leisure promo7 promo13 dpm12
county return mensware flatware homeacc lamps
linens blankets towels outdoor coats wcoat
wappar hhappar jewelry custdate numkids travtime job;
class purchase(desc) marital ntitle gender telind
aprtmnt snglmom mobile kitchen luxury dishes tmktord
statecod race origin heat numcars edlevel;
run;
The PROC DMINE statement invokes the procedure. The DATA= option identifies
the DMDB encoded training data set that is used to fit the model. The DMDBCAT=
option identifies the DMDB training data catalog.
proc dmine data=dmbexa1 dmdbcat=catexa1
The MINR2= option specifies a lower bound for the individual R-square
value to be eligible for the model selection process. Variables with R2 values
less than the MINR2 cutoff are not entered into the model. The STOPR2 specifies
a lower value for the incremental model R-square value at which the forward
selection process is stopped.
minr2=0.020 stopr2=0.0050
The NOAOV16 option prevents the DMINE procedure from including the AOV16
variables in the final forward stepwise selection process.
noaov16
The NOINTER option prevents the use of two-way interactions between
categories of class variables in the selection process.
nointer;
The VAR statement lists the numeric and categorical inputs (independent
variables).
var income homeval frequent recency age
domestic apparel leisure promo7 promo13 dpm12
county return mensware flatware homeacc lamps
linens blankets towels outdoor coats wcoat
wappar hhappar jewelry custdate numkids travtime job
marital ntitle gender telind aprtmnt snglmom mobile
kitchen luxury dishes tmktord statecod race origin heat
numcars edlevel;
The TARGET statement defines the target (response) variable.
target amount;
title 'DMINE: Continuous Target';
run;
The DMINE Procedure
Example 2: Including the AOV16 and Grouping Variables
into the Analysis (Detailed Selection Settings)
Features:
Omitting the NOAOV16 option to include the AOV16 variables into the analysis.
q
Specifying the USEGROUPS option to include only the group variables in the final model. If the
original class variable can be reduced into a group variable that contains fewer levels, then only
the group variable is considered in the final model.
q
Omitting the NOINTER option to include the two-way class interactions into the analysis.
q
Specifying the NOMONITOR option to suppress the monitor window, which displays the
execution time of the procedure.
q
This example expands on the previous example by including the AOV16, GROUP, and two-way class interaction
effects into the final forward stepwise analysis. Including these effects into the analysis may produce a better model,
but it will also increase the execution time of the DMINE procedure.
If you have not already done so, you should submit the PROC DMDB step from Example 1 before you submit the
example PROC DMINE step.
Program
proc dmine data=WORK.dmbexa1 dmdbcat=catexa1
minr2=0.020 stopr2=0.0050
usegroups
nomonitor;
var income homeval frequent recency age
domestic apparel leisure promo7 promo13 dpm12
county return mensware flatware homeacc lamps
linens blankets towels outdoor coats wcoat
wappar hhappar jewelry custdate numkids travtime job
marital ntitle gender telind aprtmnt snglmom mobile
kitchen luxury dishes tmktord statecod race origin heat
numcars edlevel;
target amount;
title 'DMINE: Continuous Target';
title2 'Add AOV16, GROUP, and 2-Way Interactions Effects';
run;
Output
Partial Listing of the R-Squares for the Target Variable
DMINE: Continuous Target
Add AOV16, GROUP, and 2-Way Interactions Effects
R-Squares for Target variable: AMOUNT
Effect DF R2
-----------------------------------------------------------
Var: LEISURE 1 0.4827
AOV16: APPAREL 12 0.4762
Class: KITCHEN*STATECOD 197 0.4271
AOV16: LEISURE 8 0.4268
Group: KITCHEN*STATECOD 9 0.4210
Class: KITCHEN*LUXURY 16 0.4019
Var: APPAREL 1 0.4001
Group: KITCHEN*LUXURY 5 0.3959
AOV16: DOMESTIC 15 0.3869
Var: DOMESTIC 1 0.3652
AOV16: FREQUENT 11 0.3418
Class: LUXURY*STATECOD 101 0.3335
Group: LUXURY*STATECOD 6 0.3284
Class: LUXURY*TMKTORD 7 0.3212
Group: LUXURY*TMKTORD 4 0.3177
Class: KITCHEN*DISHES 42 0.3128
Class: MARITAL*KITCHEN 17 0.3084
Group: KITCHEN*DISHES 5 0.3066
Var: FREQUENT 1 0.3048
Class: TMKTORD*STATECOD 110 0.3046
Group: MARITAL*KITCHEN 5 0.3033
Group: TMKTORD*STATECOD 8 0.2995
Class: KITCHEN*TMKTORD 26 0.2921
Group: KITCHEN*TMKTORD 5 0.2870
AOV16: DPM12 12 0.2770
Class: NTITLE*KITCHEN 28 0.2717
Group: NTITLE*KITCHEN 6 0.2690
Class: KITCHEN*RACE 25 0.2594
Class: KITCHEN*EDLEVEL 26 0.2572
Group: KITCHEN*RACE 5 0.2568
Class: LUXURY*DISHES 15 0.2547
Additional effects are not listed
DMINE: Continuous Target
Add AOV16, GROUP, and 2-Way Interactions Effects
R-Squares for Target variable: AMOUNT
Effect DF R2
Dostları ilə paylaş: |