The GLIMMIX Procedure
-
Overview
-
Getting Started
-
SyntaxPROC GLIMMIX StatementBY StatementCLASS StatementCODE StatementCONTRAST StatementCOVTEST StatementEFFECT StatementESTIMATE StatementFREQ StatementID StatementLSMEANS StatementLSMESTIMATE StatementMODEL StatementNLOPTIONS StatementOUTPUT StatementPARMS StatementRANDOM StatementSLICE StatementSTORE StatementWEIGHT StatementProgramming StatementsUser-Defined Link or Variance Function
-
DetailsGeneralized Linear Models TheoryGeneralized Linear Mixed Models TheoryGLM Mode or GLMM ModeStatistical Inference for Covariance ParametersDegrees of Freedom MethodsEmpirical Covariance ("Sandwich") EstimatorsExploring and Comparing Covariance MatricesProcessing by SubjectsRadial Smoothing Based on Mixed ModelsOdds and Odds Ratio EstimationParameterization of Generalized Linear Mixed ModelsResponse-Level Ordering and ReferencingComparing the GLIMMIX and MIXED ProceduresSingly or Doubly Iterative FittingDefault Estimation TechniquesDefault OutputNotes on Output StatisticsODS Table NamesODS Graphics
-
ExamplesBinomial Counts in Randomized BlocksMating Experiment with Crossed Random EffectsSmoothing Disease Rates; Standardized Mortality RatiosQuasi-likelihood Estimation for Proportions with Unknown DistributionJoint Modeling of Binary and Count DataRadial Smoothing of Repeated Measures DataIsotonic Contrasts for Ordered AlternativesAdjusted Covariance Matrices of Fixed EffectsTesting Equality of Covariance and Correlation MatricesMultiple Trends Correspond to Multiple Extrema in Profile LikelihoodsMaximum Likelihood in Proportional Odds Model with Random EffectsFitting a Marginal (GEE-Type) ModelResponse Surface Comparisons with Multiplicity AdjustmentsGeneralized Poisson Mixed Model for Overdispersed Count DataComparing Multiple B-SplinesDiallel Experiment with Multimember Random EffectsLinear Inference Based on Summary DataWeighted Multilevel Model for Survey Data
- References
The "Optimization Information" table displays important details about the optimization process.
The optimization technique that is displayed in the table is the technique that applies to any single optimization. For singly iterative methods that is the optimization method.
The number of parameters that are updated in the optimization equals the number of parameters in this table minus the number of equality constraints. The number of constraints is displayed if you fix covariance parameters with the HOLD= option in the PARMS statement. The GLIMMIX procedure also lists the number of upper and lower boundary constraints. Note that the procedure might impose boundary constraints for certain parameters, such as variance components and correlation parameters. Covariance parameters for which a HOLD= was issued have an upper and lower boundary equal to the parameter value.
If a residual scale parameter is profiled from the optimization, it is also shown in the "Optimization Information" table.
In a GLMM for which the parameters are estimated by one of the linearization methods, you need to initiate the process of computing the pseudo-response. This can be done based on existing estimates of the fixed effects, or by using the data themselves—possibly after some suitable adjustment—as an estimate of the initial mean. The default in PROC GLIMMIX is to use the data themselves to derive initial estimates of the mean function and to construct the pseudo-data. The "Optimization Information" table shows how the pseudo-data are determined initially. Note that this issue is separate from the determination of starting values for the covariance parameters. These are computed as minimum variance quadratic unbiased estimates (with 0 priors, MIVQUE0; Goodnight 1978a) or obtained from the value-list in the PARMS statement.
The ODS name of the table is OptInfo.