Previous Page | Next Page

The MDC Procedure

Functional Summary

Table 17.2 summarizes the statements and options used with the MDC procedure.

Table 17.2 MDC Functional Summary

Description

Statement

Option

Data Set Options

   

Formats the data for use by PROC MDC

MDCDATA

 

Specifies the input data set

MDC

DATA=

Specifies the output data set for CLASS STATEMENT

CLASS

OUT =

Writes parameter estimates to an output data set

MDC

OUTEST=

Includes covariances in the OUTEST= data set

MDC

COVOUT

Writes linear predictors and predicted probabilities to an output data set

OUTPUT

OUT=

Declaring the Role of Variables

   

Specifies the ID variable

ID

 

Specifies BY-group processing variables

BY

 

Printing Control Options

   

Requests all printing options

MODEL

ALL

Displays correlation matrix of the estimates

MODEL

CORRB

Displays covariance matrix of the estimates

MODEL

COVB

Displays detailed information about optimization iterations

MODEL

ITPRINT

Suppresses all displayed output

MODEL

NOPRINT

Model Estimation Options

   

Specifies the choice variables

MODEL

CHOICE=()

Specifies the convergence criterion

MODEL

CONVERGE=

Specifies the type of covariance matrix

MODEL

COVEST=

Specifies the starting point of the Halton sequence

MODEL

HALTONSTART=

Specifies options specific to the HEV model

MODEL

HEV=()

Sets the initial values of parameters used by the iterative optimization algorithm

MODEL

INITIAL=()

Specifies the maximum number of iterations

MODEL

MAXITER=

Specifies the options specific to mixed logit

MODEL

MIXED=()

Specifies the number of choices for each person

MODEL

NCHOICE=

Specifies the number of simulations

MODEL

NSIMUL=

Specifies the optimization technique

MODEL

OPTMETHOD=

Specifies the type of random number generators

MODEL

RANDNUM=

Specifies that initial values are generated using random numbers

MODEL

RANDINIT

Specifies the rank dependent variable

MODEL

RANK

Specifies optimization restart options

MODEL

RESTART=()

Specifies a restriction on inclusive parameters

MODEL

SAMESCALE

Specifies a seed for pseudo-random number generation

MODEL

SEED=

Specifies a stated preference data restriction on inclusive parameters

MODEL

SPSCALE

Specifies the type of the model

MODEL

TYPE=

Specifies normalization restrictions on multinomial probit error variances

MODEL

UNITVARIANCE=()

Controlling the Optimization Process

   

Specifies upper and lower bounds for the parameter estimates

BOUNDS

 

Specifies linear restrictions on the parameter estimates

RESTRICT

 

Specifies nonlinear optimization options

NLOPTIONS

 

Nested Logit Related Options

   

Specifies the tree structure

NEST

LEVEL()=

Specifies the type of utility function

UTILITY

U()=

Output Control Options

   

Outputs predicted probabilities

OUTPUT

P=

outputs estimated linear predictor

OUTPUT

XBETA=

Test Request Options

   

Requests Wald, Lagrange multiplier, and likelihood ratio tests

TEST

ALL

Requests the Wald test

TEST

WALD

Requests the Lagrange multiplier test

TEST

LM

Requests the likelihood ratio test

TEST

LR

Previous Page | Next Page | Top of Page