| Chapter 52 TheMCMCProcedure- J# _% X) a, x. P; S* Q9 I Contents0 e L: {/ }1 ~4 N! w& o. ? Overview:MCMCProcedure ............................ 3478 PROCMCMCComparedwithOtherSASProcedures ............3479 GettingStarted:MCMCProcedure .......................... 3479 SimpleLinearRegression ...........................34804 s! z5 \: S. L TheBehrens-FisherProblem ..........................3488 Mixed-EffectsModel .............................3492- P/ z5 m# x9 \( F; ^8 [4 E4 C Syntax:MCMCProcedure .............................. 3495 PROCMCMCStatement ...........................3496 ARRAYStatement ...............................3508 BEGINCNST/ENDCNSTStatement .....................3509! ^# ~9 o# h, s BEGINNODATA/ENDNODATAStatements .................3511 BYStatement .................................35119 A5 g+ H3 n) Y MODELStatement ...............................3512' C" b$ C! Y7 a* x, z* [( S PARMSStatement ...............................3515 PRIOR/HYPERPRIORStatement .......................35162 C w3 O* c+ ~1 L ProgrammingStatements ...........................3516 UDSStatement .................................3518 Details:MCMCProcedure .............................. 3522" L/ Z* G3 `0 r" \0 ^' P6 x# i: a7 ` HowPROCMCMCWorks ..........................3522 BlockingofParameters ............................3523 Samplers ....................................3524 TuningtheProposalDistribution .......................3525. ]6 U; T) o" h5 k; y9 N InitialValuesoftheMarkovChains ......................3528 AssignmentsofParameters ..........................3528 StandardDistributions .............................35301 Q$ M- I. X1 S SpecifyingaNewDistribution .........................3541 UsingDensityFunctionsintheProgrammingStatements ...........35421 P1 Q0 r/ f! s4 k0 p TruncationandCensoring ...........................3544 MultivariateDensityFunctions ........................35463 ]4 E1 V$ w) H! n SomeUsefulSASFunctions ..........................3549 MatrixFunctionsinPROCMCMC ......................3551' I# ^* T' C- k: l2 \5 A ModelingJointLikelihood ...........................3556& i0 X3 l2 b2 G* R( s: R$ t$ L! Z RegeneratingDiagnosticsPlots ........................3557/ G: w: ?2 r8 x: [ PosteriorPredictiveDistribution ........................35606 e( m8 w# @) x s+ e. _/ l, h8 a3 y HandlingofMissingData ...........................3565 FloatingPointErrorsandOverflows ......................3565, ^# P- G9 b7 d; B HandlingErrorMessages ...........................3568 ComputationalResources ...........................3570 DisplayedOutput ................................3571; c# c0 z" j4 w/ c# ] k. p) ~. h8 b ODSTableNames ...............................3575 ODSGraphics .................................3577 Examples:MCMCProcedure ............................ 3578 S; Y. g4 n8 a1 M" e F/ W! O- w Example52.1:SimulatingSamplesFromaKnownDensity .........3578 Example52.2:Box-CoxTransformation ...................3583 Example52.3:GeneralizedLinearModels ..................3592 Example52.4:NonlinearPoissonRegressionModels ............3605- E. ?, t) d6 L. C! ^# c, X# g Example52.5:Random-EffectsModels ...................3614 Example52.6:ChangePointModels .....................3630% k; q0 t) R9 s$ V4 s: J5 ] Example52.7:ExponentialandWeibullSurvivalAnalysis ..........3634 Example52.8:CoxModels ..........................3647 Example52.9:NormalRegressionwithIntervalCensoring .........3664$ j/ S+ V9 m0 c6 R8 K( k' ~% ]1 \ Example52.10:ConstrainedAnalysis ....................3666 Example52.11:ImplementaNewSamplingAlgorithm ...........3672 Example52.12:UsingaTransformationtoImproveMixing .........3683 Example52.13:Gelman-RubinDiagnostics .................3693/ _' w4 L3 M. [$ y: D3 k8 [ 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 |