\newpage
Subgroup analysis are only exploratory unless such analyses were pre-specified in a study protocol at the design stage. The users should be cautious with the choices of the priors in Bayesian subgroup analysis. The users are encouraged to conduct sensitivity analysis by trying different parameter values in the prior distributions and examining the robustness of the results.
dat.all <- get.data() summary(dat.all)
subgrp <- get.subgrp() pander::pandoc.table(subgrp)
arst <- ana.rst() for (myi in 1:length(arst)) { cat("## Model:", names(arst)[myi], "\n\n"); cat("### Subgroup Effects\n\n"); pander::pandoc.table(bzSummary(arst[[myi]], sel.grps=input$subgrpselected)); cat("\n\n"); bzPlot(arst[[myi]], sel.grps=input$subgrpselected); cat("\n\n"); bzForest(arst[[myi]], sel.grps=input$subgrpselected); cat("\n\n"); cat("### Subgroup Effects Comparison\n\n"); pander::pandoc.table(bzSummaryComp(arst[[myi]], sel.grps=input$subgrpselected)); cat("\n\n"); bzPlotComp(arst[[myi]], sel.grps=input$subgrpselected); cat("\n\n"); bzForestComp(arst[[myi]], sel.grps=input$subgrpselected); cat("\n\n"); }
dat <- get.subgrp(); disp.cut <- input$displaycut; disp.digit <- input$displaydigit; sub.cov <- get.sub.cov(); dic <- get.all.dic(); min.mdl <- which.min(dic); mdls <- get.ana.models(); rst <- bzRptTbl(arst, dat, sub.cov, disp.cut, disp.digit);
Based on leave-one-out cross-validation information
criterion (LOOIC), results from the r mdls[min.mdl]
model are reported here. However, this does not imply that the selected
model is significantly better than the other models. Please consider the
totality of the information for drawing the conclusions.
Based on LOOIC, r mdls[min.mdl]
model is selected for the analysis dataset.
The summary result table is as follows:
pander::pandoc.table(rst); cat("\n\n");
Any scripts or data that you put into this service are public.
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.