BAYES REGRESSION

BAYES REGRESSION 在 自訂表格及進階統計量中可用。

BAYES REGRESSION dependent varname [BY factor list] [WITH covariate list]
     /CRITERIA
          CILEVEL = {95**} {value}
          TOL = {0.000001**} {value}
          MAXITER = {2000**} {integer}
     /DESIGN
          REGWGT = {variable}@@
          FACTORS = {factors}##
          COVARIATES = {covariates}$$
     /INFERENCE
          ANALYSIS = {POSTERIOR**} {BAYESFACTOR} {BOTH}
     /ESTBF
          COMPUTATION = {JZS**} {ZELLNER(value)} {HYPER(3** {value})} {ROUDER(1** {value})}
          COMPARE = {NULL**} {FULL}
     /PRIOR
          TYPE = {REFERENCE**} {CONJUGATE}
          SHAPEPARAM = {value}
          SCALEPARAM = {value}
          MEANVECTOR = {values}
          VMATRIX = {IDENTITY**} {lower_triangular_matrix}%%
     [/PREDICT
          REGRESSORS = {values}]
     [/SAVE
          [ MEANPRED[(varname)]] [VARIANCE[(varname)]] [MODE[(varname)]] [CIMEANPREDL[(varname)]]
          [ CIMEANPREDU[(varname)]]]
     [/OUTFILE
          PARAMETER = 'filename']
     [/PLOT
          COVARIATES = {covariate_sublist}
          FACTORS = {factor_sublist}
          MAX CATEGORY = {2**} {integer}^^
          INTERCEPT = {FALSE**} {TRUE}
          ERRORVAR = {FALSE**} {TRUE}
          BAYESPRED = {FALSE**} {TRUE}]&&
     [/FTEST ⊕
          [ LABEL = 'labelName']
          VARIABLES = {varlist}⊖
          INTERCEPT = {FALSE**} {TRUE}⊗
          [ TESTVALUES = {values}]] ⊜

**如果省略了次指令或關鍵字，則為預設值。

@ @ 只容許一個迴歸加權變數。

## 必須是因素清單的子集。

$$ 必須是共變數清單的子集。

%% 以分號區隔列。

^ ^ 控制項要繪製的因子水準數目。

&& Can be TRUE when values are specified for REGRESSORS.

容許多個 /FTEST 次指令。

-必須是 COVARIATES 和 FACTS 的子集。

當 VARIABLES 是空的時，其必須為 TRUE。

如果未指定，則會假設所有值都等於 0。

發行歷程