LukasWallrich / metaUI

Create shiny apps that allow explore of meta-analysis results
https://lukaswallrich.github.io/metaUI/
GNU General Public License v3.0
6 stars 0 forks source link

Ensure forest plot limitaton works on shinyapps.io #30

Open LukasWallrich opened 7 months ago

LukasWallrich commented 7 months ago

The validate(need()) code to limit the number of effects shown in a forest plot works locally, but struggles on shinyapps.io - there, something more complex may be needed, along these lines:

output$foreststudies <- renderPlot({
    tryCatch({
      df <- df_filtered()
      #TK: consider different forest plot package that is easier to handle
      if(nrow(df) > 250) {
        stop("Forest plots can only be displayed here with 250 effect sizes or fewer. Please filter the dataset further before proceeding.")
      }

      rve <- robumeta::robu(metaUI__effect_size ~ 1, data = df, studynum = metaUI__study_id, var.eff.size = metaUI__variance, small = FALSE)

      robumeta::forest.robu(rve, es.lab = "metaUI__es_label", study.lab = "metaUI__study_id", "Effect size" = metaUI__effect_size)
    }, error = function(e) {
      # This will stop the execution and display the error message in the plot output
      plot(NA, xlim = c(0, 1), ylim = c(0, 1), type = "n", xlab = "", ylab = "", axes = FALSE)
      text(0.5, 0.5, paste(e$message), cex = 1.2, col = "red")
    })
  }, height = function() {
    # Safeguard in case df_filtered() is NULL or has an unexpected error
    df <- tryCatch({
      df_filtered()
    }, error = function(e) {
      return(NULL)
    })
    if(is.null(df) || nrow(df) > 250) {
      return(200)  # Default height if df is NULL
    } else {
      return(400 + 25 * nrow(df_filtered()))
    }
  }, width = 900)