Dashboard speed optimization by reducing redundant plot function calls

[rafaelcatoia]

Because of how reactivity in shiny works, there were many redundant calls.

Every time that we update a reactive variable, it automatically triggers functions that depends on this reactive variable. For example: changing input$targetVariable triggers summaryPlot, so there is no need

In this PR we removed some of those forced calls of functions that would already be triggered by reactivity.

For example, when changing colours, there is no need to call summaryPlot since it will alread be triggered by the change of the reactive variable COLOR_SEED()

Here is an example of change that was made:

From:

  observeEvent(input$refreshColors, {
    COLOR_SEED(floor(runif(1, 1, 1000)))
    output$summaryPlot <- renderPlotly({
      summaryPlot()
    })
  })

to:

  observeEvent(input$refreshColors,
    {
      COLOR_SEED(floor(runif(1, 1, 1000)))
      USE_CURR_TRUTH <<- TRUE
    },
    ignoreInit = TRUE
  )

[nmdefries]

Initial review with general comments. I didn't look at the logic in-depth yet and haven't tried running the dashboard myself. I'll plan to do a local build in docker to make sure the production version will work.

[nmdefries]

All the code changes look good and everything works in docker!

Before I merge this, could you describe in the PR what general changes you made? In the box that says "No description provided."