| Chapter 52 TheMCMCProcedure6 y& m4 D4 E# H9 L/ ] Contents Overview:MCMCProcedure ............................ 3478 PROCMCMCComparedwithOtherSASProcedures ............3479 GettingStarted:MCMCProcedure .......................... 3479 SimpleLinearRegression ...........................34802 \% q- V6 `7 m0 G2 | TheBehrens-FisherProblem ..........................3488 Mixed-EffectsModel .............................3492 Syntax:MCMCProcedure .............................. 34959 Q- x; v7 w& N PROCMCMCStatement ...........................3496 ARRAYStatement ...............................3508% K' u3 \' y& f( O0 t( t( t5 i1 \$ I BEGINCNST/ENDCNSTStatement .....................3509 BEGINNODATA/ENDNODATAStatements .................35111 l# _# C. i q8 m; y BYStatement .................................35119 |. P; q( X) c) k5 N1 J MODELStatement ...............................3512 PARMSStatement ...............................3515 PRIOR/HYPERPRIORStatement .......................3516: W7 d/ |" ]4 p8 I" S4 U ProgrammingStatements ...........................3516 UDSStatement .................................3518+ c; S0 g# D! ^/ ^( ~ Details:MCMCProcedure .............................. 3522 HowPROCMCMCWorks ..........................3522! S4 b' B. m' X/ Q+ W BlockingofParameters ............................35235 P* u+ h# Y1 t Samplers ....................................3524 TuningtheProposalDistribution .......................3525 InitialValuesoftheMarkovChains ......................3528% T }* P; w5 d8 i AssignmentsofParameters ..........................3528 StandardDistributions .............................3530: M- c' o4 C$ ^: |) T1 }! @ SpecifyingaNewDistribution .........................3541 UsingDensityFunctionsintheProgrammingStatements ...........3542 TruncationandCensoring ...........................3544 MultivariateDensityFunctions ........................3546 SomeUsefulSASFunctions ..........................3549 MatrixFunctionsinPROCMCMC ......................3551- z5 |# z& O$ F! f ModelingJointLikelihood ...........................3556 RegeneratingDiagnosticsPlots ........................3557' ~. j+ ^: e" s4 } PosteriorPredictiveDistribution ........................35601 k8 x V" n, o/ N* X/ r HandlingofMissingData ...........................35659 y# ~/ q# f- P" s; Y/ v5 ?$ ?6 ] FloatingPointErrorsandOverflows ......................35655 E# i. c& c: O' m' D( R4 y5 L HandlingErrorMessages ...........................3568; h. L2 P6 \5 T! _9 J: R ComputationalResources ...........................3570 DisplayedOutput ................................3571 ODSTableNames ...............................3575 ODSGraphics .................................3577 Examples:MCMCProcedure ............................ 3578% m7 N9 F; l. }+ d, y) U* } Example52.1:SimulatingSamplesFromaKnownDensity .........3578 Example52.2:Box-CoxTransformation ...................35832 w4 l9 V& X' |$ N; V Example52.3:GeneralizedLinearModels ..................3592' u, a: w: W. G+ v- n6 L9 ^ Example52.4:NonlinearPoissonRegressionModels ............3605. p) v8 B7 P4 @/ R1 E7 m ^2 n3 O1 R Example52.5:Random-EffectsModels ...................36148 m6 d/ ]6 r L/ e Example52.6:ChangePointModels .....................3630 Example52.7:ExponentialandWeibullSurvivalAnalysis ..........3634 Example52.8:CoxModels ..........................3647 Example52.9:NormalRegressionwithIntervalCensoring .........3664, C6 ]4 W7 }) K1 i5 h Example52.10:ConstrainedAnalysis ....................3666 Example52.11:ImplementaNewSamplingAlgorithm ...........3672 Example52.12:UsingaTransformationtoImproveMixing .........3683$ c* {$ D. Q1 @ Example52.13:Gelman-RubinDiagnostics .................3693 References ...................................... 3700 |
| 欢迎光临 数学建模社区-数学中国 (http://www.madio.net/) | Powered by Discuz! X2.5 |