The DMREG Procedure
Overview
DMREG enables you to fit both linear and logistic regression models. Linear regression attempts to
predict the value of a continuous target as a linear function of one or more independent inputs. Logistic
regression attempts to predict the probability that a categorical (binary, ordinal, or nominal) target will
acquire the event of interest as a function of one or more independent inputs. The procedure supports
forward, backward, and stepwise selection methods. It also allows you to score data sets or generate SAS
DATA step code to score a data set.
Copyright 2000 by SAS Institute Inc., Cary, NC, USA. All rights reserved.
The DMREG Procedure
Procedure Syntax
PROC DMREG < option(s)>;
MODEL dependent=independent(s) < / model-option(s)>;
CLASS variable(s);
CODE code-option(s);
DECISION DECDATA=< libref.> SAS-data-set< DECVARS=decision-variable(s)>
< option(s)>;
FREQ variable;
NLOPTIONS nonlinear-option(s);
REMOTE remote-option(s);
SCORE scoring-option(s);
Copyright 2000 by SAS Institute Inc., Cary, NC, USA. All rights reserved.
The DMREG Procedure
PROC DMREG Statement
Invokes the DMREG procedure.
PROC DMREG < option(s)>;
Required Arguments
DATA= SAS-data-set
Identifies the training data set.
DMDBCAT= SAS-catalog
Identifies the training data catalog.
Options
COVOUT
Specifies that the OUTEST= data set is to include the variance-covariance matrix of the parameter
estimates.
DESCENDING
Specifies that the order of categorical target is to be reversed.
ESTITER=n
Specifies that the OUTEST= data set contains parameter estimates and fit statistics (for training,
test, and validation data) for every nth iteration.
Default:
0. Only the parameter estimates of the final iteration are output.
INEST= SAS-data-set
Identifies the data set that contains initial estimates.
MINIMAL
Specifies the use of minimal resources to fit a logistic regression model. Memory for the Hessian
matrix is not needed. The optimization defaults to the conjugate gradient technique and standard
errors of the regression parameters are not computed. Model selection is disabled when this option
is specified. This option does not apply to the normal error regression models.
NAMELEN=n
Specifies the length of effect names in the printed output to be n characters, where n is a value
between 20 and 200. The default length is 20 characters.
OUTEST= SAS-data-set
Identifies the output data set containing estimates and fit statistics. See for more information.
NOPRINT
Suppresses all printed output.
SIMPLE
Prints simple descriptive statistics of the input variables.
TESTDATA= SAS-data-set
Identifies the data set containing test data.
VALIDATA= SAS-data-set
Identifies the data set containing validation data.
Copyright 2000 by SAS Institute Inc., Cary, NC, USA. All rights reserved.
The DMREG Procedure
CLASS Statement
Specifies one or more categorical variables to be used in the analysis.
CLASS variable(s);
Required Argument
variable(s)
Specifies a list of categorical variables to be used in the analysis. You must specify the target
variable if it has a categorical (binary, ordinal, or nominal) measurement level.
Copyright 2000 by SAS Institute Inc., Cary, NC, USA. All rights reserved.
The DMREG Procedure
CODE Statement
Controls the creation of SAS code that can be used to score data sets.
Tip: If neither FILE= nor METABASE= is specified, then the SAS code is written to the SAS log.
CODE < code-option(s)>;
CODE Options
ERROR
Specifies that the error function is to be computed.
FILE=
Specifies the path for writing the code to an external file. For example,
FILE="c:\mydir\scorecode.sas".
FORMAT=
Specifies numeric formats for printing the estimated parameters.
GROUP=
Specifies the group identifier (up to four characters) for group processing.
METABASE=mylib.mycat.myentry
Specifies the code catalog entry to which the results are written.
RESIDUAL
Specifies that residuals are to be computed.
Copyright 2000 by SAS Institute Inc., Cary, NC, USA. All rights reserved.
The DMREG Procedure
DECISION Statement
Specifies information used for decision processing in the DECIDE, DMREG, NEURAL, and
SPLIT procedures. This documentation applies to all four procedures.
Tip: The DECISION statement is required for the DMREG and NEURAL procedures. It is optional
for PROC SPLIT.
DECISION DECDATA= SAS-data-set < DECVARS=decision-variable(s)>< option(s)>;
DECDATA= SAS-data-set
Specifies the input data set that contains the decision matrix. The DECDATA= data set must
contain the target variable.
Note: The DECDATA= data set may also contain decision variables specified by means of the
DECVARS= option, and prior probability variable(s) specified by means of the PRIORVAR=
option.
The target variable is specified by means of the TARGET statement in the DECIDE, NEURAL,
and SPLIT procedures or by using the MODEL statement in the DMREG procedure. If the target
variable in the DATA= data set is categorical then the target variable of the DECDATA= data set
should contain the category values, and the decision variables will contain the common
consequences of making those decisions for the corresponding target level. If the target variable is
interval, then each decision variable will contain the value of the consequence for that decision at
a point specified in the target variable. The unspecified regions of the decision function are
interpolated by a piecewise linear spline.
Tip:
The DECDATA= data set may be of TYPE=LOSS, PROFIT, or REVENUE.
If unspecified, TYPE=PROFIT is assumed by default. TYPE= is a data set
option that should be specified when the data set is created.
DECVARS=decision-variable(s)
Specifies the decision variables in the DECDATA= data set that contain the target-specific
consequences for each decision.
Default:
None
COST=cost-option(s)
Specifies numeric constants that gives the cost of a decision, or variables in the DATA= data set
that contain the case-specific costs, or any combination of constants and variables. There must be
the same number of cost constants and variables as there are decision variables in the DECVARS=
Dostları ilə paylaş: |