DYNAMICLINEAR Procedure

Functional Summary

The statements and options available in the DYNAMICLINEAR procedure are summarized in Table 1.

Table 1: Functional Summary

Description Statement Option
Input Data Table Options
Specifies the input data table PROC DYNAMICLINEAR DATA=
Specifies the initial values for all discount factors DISCOUNTFACTOR INDISCOUNTFACTOR=
Specifies the initial covariance values by using a row-oriented table INITIALDISTRIBUTION INCOVARIANCE=
Specifies the initial variance values (diagonal) for the initial distribution’s covariances INITIALDISTRIBUTION INCOVARIANCE(DIAGONAL)=
Specifies the initial values for the shape and rate parameters of the normal-gamma distribution INITIALDISTRIBUTION INGAMMA=
Specifies the initial values of means by using a column-oriented table INITIALDISTRIBUTION INMEAN=
Specifies the adjacency matrix of the parental sets PARENTALSET INADJACENCY=
Output Data Table Options
Specifies the table to use for validating the input discount factors DISCOUNTFACTOR OUTVERIFICATION=
Specifies the output table for the covariances of prior/naive posterior/posterior normal-gamma distributions FILTER COVARIANCE=
Specifies the output table for the shape and rate parameters of prior/naive posterior/posterior normal-gamma distributions FILTER GAMMA=
Specifies the output table for Kullback-Leibler divergence information in the decoupling step of updating the distribution from naive posterior to posterior FILTER KLDIVERGENCE=
Specifies the output table for raw random samples of precisions during forecasting/importance sampling FILTER LAMBDA=
Specifies the output table for the means of prior/naive posterior/posterior normal-gamma distributions FILTER MEAN=
Specifies the output table for raw random samples of observation errors during forecasting FILTER NU=
Writes in-sample forecasting results and statistics to an output table FILTER OUT=
Writes the covariances of samples of in-sample forecasting observations to an output table FILTER OUTCOVARIANCE=
Writes samples of in-sample forecasting observations to an output table FILTER OUTSIM=
Specifies the output table for raw random samples of state vectors during forecasting/importance sampling FILTER THETA=
Writes out-of-sample forecasting results and statistics to an output table FORECAST OUT=
Writes the covariances of samples of out-of-sample forecasting observations to an output table FORECAST OUTCOVARIANCE=
Writes samples of out-of-sample forecasting observations to an output table FORECAST OUTSIM=
Specifies the output table for the covariances of the initial distribution INITIALDISTRIBUTION COVARIANCE=
Specifies the output table for the shape and rate parameters of the initial distribution INITIALDISTRIBUTION GAMMA=
Specifies the output table for the means of the initial distribution INITIALDISTRIBUTION MEAN=
Specifies the output table to use for verifying the input adjacency matrices PARENTALSET OUTVERIFICATION=
Common Options
Specifies the number of simulations in filtering and forecasting PROC DYNAMICLINEAR NSIMULATIONS=
Specifies a nonnegative integer to use as the seed for generating random number sequences PROC DYNAMICLINEAR SEED=
ID Variable
Identifies observations in the input data table by specifying a variable for the time series data ID
Variable and Parameter Declaration Options
Specifies the model’s stochastic volatility discount factors DISCOUNTFACTOR BETA=
Specifies the model’s discount factors for regressors in the parental sets DISCOUNTFACTOR DELTAGAMMA=
Specifies the model’s discount factors for regressors not in the parental sets DISCOUNTFACTOR DELTAPHI=
Specifies the starting covariance settings for the initial distribution INITIALDISTRIBUTION COVARIANCE=
Specifies the default initial value for all means INITIALDISTRIBUTION MEAN=
Specifies the starting setting for the initial distribution’s rate parameter INITIALDISTRIBUTION RATE=
Specifies the starting setting for the initial distribution’s shape parameter INITIALDISTRIBUTION SHAPE=
Specifies the variables to use for the model MODEL
Inference Options
Specifies the number of observations before the end of the data at which the forecasts begin FILTER BACK=
Specifies the number of the multistep forecast values to compute during filtering FILTER LEAD=
Specifies which parameters of the naive posterior distribution to output FILTER NAIVEPOSTERIOR=
Specifies which filtering information to write to output tables FILTER OUTFILTERINFO=
Specifies which parameters of the posterior distribution to output FILTER POSTERIOR=
Specifies which raw random samples to output during forecasting FILTER PREDICTIONSAMPLE=
Specifies which parameters of the prior distribution to output FILTER PRIOR=
Specifies which raw random samples to output during the recoupling of importance sampling FILTER RECOUPLINGSAMPLE=
Specifies the number of the multistep forecast values to compute during forecasting FORECAST LEAD=
GPU Options
Specifies that the model perform calculations by using GPU hardware GPU
Specifies the computation precision that the model uses for calculations on GPU hardware GPU PRECISION=


Last updated: July 09, 2026