TURF - Upset Plot

Creates an upset plot, showing the intersection between binary variables. It is useful for understanding the results of a TURF analysis. Our TURF eBook contains more information on how to interpret and use this visualization.

Usage

In Displayr, go to Insert > More > TURF > Upset plot.

In Q, go to Automate > Browse Online Library > TURF > Upset plot.

To run this function, the data must consist of more than one variable in a binary format (1's and 0's), where the 1's represent selected values and 0's represent options that were not selected. It should look like the following in your raw data:

Variable sets or questions that contain binary data are represented with an icon of either two or four check-boxes.

In the object inspector on the right of the screen, under Inputs > Alternatives select one or more binary variable sets (questions), change any other settings as required.

Options

Binary variables Two or more binary variables or a question of binary variables.

COLOR > Column color The color of the columns showing the size of the intersections.

COLOR > Dot color The color of dots that appear below the columns.

COLOR > Background shade The color of the shading behind the dots.

COLOR > Bar color The color of bars showing the size of each of the alternatives (selected values of the binary variables).

Acknowledgements

Uses the UpsetR package.

Code

var heading_text = "Upset Plot";
if (!!form.setObjectInspectorTitle)
    form.setObjectInspectorTitle(heading_text);
else 
    form.setHeading(heading_text);


form.dropBox({name: "formData",
              label: "Binary variables",
              types:["Q: pickany, pickanycompact, pickanygrid",
                     "V: numeric, categorical, ordered categorical"],
              multi: true,
              required: true})
form.group("COLOR")
form.colorPicker({name: "formMainBarColor",
                  label: "Column color",
                  default_value: "#3e7dcc"})
form.colorPicker({name: "formMatrixColor",
                  label: "Dot color",
                  default_value: "#3e7dcc"})
form.colorPicker({name: "formShadeColor",
                  label: "Background shade",
                  default_value: "#8cc0ff"})
form.colorPicker({name: "formSetsBarColor",
                  label: "Bar color",
                  default_value: "#8cc0ff"})

x <- list()
for (nm in names(formData))
{
    elem <- formData[[nm]]
    if (is.data.frame(elem))
    {
        for (nm2 in setdiff(names(elem), c("NET", "SUM")))
            x[[nm2]] <- elem[[nm2]]
    }
    else
        x[[nm]] <- elem
}
x = as.data.frame(x, check.names = FALSE)
if (NCOL(x) <= 1)
    stop("A minimum of 2 variables is required to create an upset plot.")
x = x[QFilter, ]
levels = suppressWarnings(as.character(unique(unlist(lapply(x, unique)))))
if (any(!complete.cases(x)))
    stop("Data contains missing values. This is not possible for Upset plots. Please either recode them or filter the plot.")
if (!all(levels %in% as.character(0:1))) {
    types = if (get0("productName") == "Q") "Set the Question Type of the variables to Pick Any or Pick Any - Grid."
        else "Set the structure of the variable sets to Binary - Multi or Binary - Grid."
    stop("Upset plots require binary variables. ", types)}
if (!is.null(QPopulationWeight))
    stop("A weight has been applied. This is not possible for Upset plots.")
UpSetR::upset(x,
      keep.order = TRUE,
      nsets = ncol(x),
      order.by = "freq",
      mainbar.y.label = "Intersection size",
      set_size.numbers_size = 6,
      color = "orange",
      matrix.color = formMatrixColor,
      main.bar.color = formMainBarColor,
      shade.color = formShadeColor,
      shade.alpha = 0.1,
      sets.bar.color = formSetsBarColor,
      sets.x.label = "Number of observations",
      text.scale = 1.5)