SMC Procedure

Functional Summary

The statements and options available in the SMC procedure are summarized in Table 1. The sections after the table provide a description of each statement in alphabetical order, following the PROC SMC statement.

Table 1: Functional Summary

Description Statement Option
Input Data Table Options
Specifies the input data table PROC SMC DATA=
Specifies the input data table for model information SCORE INMODEL=
Output Data Table Options
Writes the model information to the specified output data set SCORE OUTMODEL=
Writes the evaluation estimates to the specified output data set EVALUATE OUT=
Writes the summary of the filtering estimates to the specified output data set FILTER OUT=
Writes the filtering estimates to the specified output data set FILTER OUTSIM=
Writes the summary of the forecasting estimates to the specified output data set FORECAST OUT=
Writes the forecasting estimates to the specified output data set FORECAST OUTSIM=
Writes the summary of the smoothing estimates to the specified output data set SMOOTH OUT=
Writes the smoothing estimates to the specified output data set SMOOTH OUTSIM=
SMC Common Options
Specifies the number of particles to use in the SMC algorithms PROC SMC NPARTICLE=
Specifies the random seed to use in the SMC algorithms PROC SMC SEED=
Variable and Parameter Declaration Options
Specifies the names and initial values of the model parameters PARAMETERS
Specifies the lower bounds of the model parameters PARAMETERS LOWER=
Specifies the upper bounds of the model parameters PARAMETERS UPPER=
Specifies the names of state variables STATEVAR
Specifies the names of dependent and independent (exogenous) variables VAR
Model Specification Options
Specifies the initialization distribution of the state variable INITSTATE
Specifies the proposal initialization distribution of the state variable INITSTATE PROPOSAL=
Specifies the transition distribution of the state variable STATE
Specifies the proposal transition distribution of the state variable STATE PROPOSAL=
Specifies the measurement distribution of the dependent variable MODEL
Specifies the auxiliary distribution of the dependent variable AUXMODEL
Specifies the name of the model of interest SUBMITMODEL NAME=
Specifies the number of parameters SUBMITMODEL NPARM=
Specifies the number of state variables SUBMITMODEL NSTATE=
Specifies the beginning of program statements SUBMITMODEL
Specifies the end of program statements ENDSUBMIT
ID Variables
Specifies the variable that identifies the time or sequence ID TIME=
Parameter Estimation Options
Specifies the prior distribution and range of the parameters PRIOR
Specifies the number of particles to use in the parameter estimation process LEARN NPARTICLE=
Specifies the particle filter algorithm to use in the parameter estimation process LEARN ALGORITHM=
Specifies the parameter estimation method and options LEARN METHOD=PMCMC
Inference Options
Specifies the particle filter algorithm to use to solve the evaluation problem EVALUATE ALGORITHM=
Specifies the number of particles to use in the evaluation process EVALUATE NPARTICLE=
Specifies the type of parameter values from the posterior distribution to use in the evaluation process EVALUATE PARM=
Specifies the particle filter algorithm to use to solve the filtering problem FILTER ALGORITHM=
Specifies the number of particles to use in the filtering process FILTER NPARTICLE=
Specifies the type of parameter values from the posterior distribution to use in the filtering process FILTER PARM=
Specifies the particle filter algorithm to use to get particles before the (multi-step) forecasts FORECAST ALGORITHM=
Specifies how many periods to forecast FORECAST LEAD=
Specifies the number of particles to use in the forecasting process FORECAST NPARTICLE=
Specifies the type of parameter values from the posterior distribution to use in the forecasting process FORECAST PARM=
Specifies that particles are to be resampled before forecasting FORECAST RESAMPLE
Specifies the particle filter algorithm to use in the smoothing process SMOOTH ALGORITHM=
Specifies the number of particles to use in the smoothing process SMOOTH NPARTICLE=
Specifies the smoothing method SMOOTH METHOD=
Specifies the type of parameter values from the posterior distribution to use in the smoothing process SMOOTH PARM=


Last updated: July 09, 2026