Box Plot - ScottPlot 5.0 Cookbook

This page is part of the ScottPlot 5.0 Cookbook

⚠️ ScottPlot 5.0.8-beta is a preview package

This page describes a beta release of ScottPlot. It is available on NuGet as a preview package, but its API is not stable and it is not recommended for production use. See the ScottPlot Versions page for more information.

Box Plot Quickstart

Box plots can be added from a series of values.

ScottPlot.Plot myPlot = new();

// TODO: move this functionality to the RandomDataGenerator class
Random rand = new(0);
ScottPlot.Plottables.Box CreateRandomBox()
{
    int N = 50;
    double mean = rand.NextDouble() * 3;
    double stdDev = rand.NextDouble() * 3;

    double[] values = Generate.RandomNormal(N, mean, stdDev);
    Array.Sort(values);
    double min = values[0];
    double q1 = values[N / 4];
    double median = values[N / 2];
    double q3 = values[3 * N / 4];
    double max = values[N - 1];

    return new ScottPlot.Plottables.Box
    {
        WhiskerMin = min,
        BoxMin = q1,
        BoxMiddle = median,
        BoxMax = q3,
        WhiskerMax = max,
    };
}

// TODO: construct 3 boxes manually with hard-coded numerical values for simplicity
List<ScottPlot.Plottables.Box> boxes = Enumerable.Range(0, 5)
    .Select(x => CreateRandomBox())
    .ToList();

myPlot.Add.Box(boxes);

myPlot.SavePng("box-plot-quickstart.png");

Individual Box Plots

One can easily create a box plot with only a single box.

ScottPlot.Plot myPlot = new();

Random rand = new(0);

ScottPlot.Plottables.Box CreateRandomBox()
{
    int N = 50;
    double mean = rand.NextDouble() * 3;
    double stdDev = rand.NextDouble() * 3;

    double[] values = Generate.RandomNormal(N, mean, stdDev);
    Array.Sort(values);
    double min = values[0];
    double q1 = values[N / 4];
    double median = values[N / 2];
    double q3 = values[3 * N / 4];
    double max = values[N - 1];

    return new ScottPlot.Plottables.Box
    {
        WhiskerMin = min,
        BoxMin = q1,
        BoxMiddle = median,
        BoxMax = q3,
        WhiskerMax = max,
    };
}

// TODO: construct 3 boxes manually with hard-coded numerical values for simplicity
List<ScottPlot.Plottables.Box> boxes = new() { CreateRandomBox() };

myPlot.Add.Box(boxes);

myPlot.SavePng("individual-box-plots.png");

Horizontal Box Plots

Box plots can be oriented horizontally, similarly to bar plots.

ScottPlot.Plot myPlot = new();

Random rand = new(0);

ScottPlot.Plottables.Box CreateRandomBox()
{
    int N = 50;
    double mean = rand.NextDouble() * 3;
    double stdDev = rand.NextDouble() * 3;

    double[] values = Generate.RandomNormal(N, mean, stdDev);
    Array.Sort(values);
    double min = values[0];
    double q1 = values[N / 4];
    double median = values[N / 2];
    double q3 = values[3 * N / 4];
    double max = values[N - 1];

    return new ScottPlot.Plottables.Box
    {
        WhiskerMin = min,
        BoxMin = q1,
        BoxMiddle = median,
        BoxMax = q3,
        WhiskerMax = max,
    };
}

// TODO: construct 3 boxes manually with hard-coded numerical values for simplicity
List<ScottPlot.Plottables.Box> boxes = Enumerable.Range(0, 5)
    .Select(x => CreateRandomBox())
    .ToList();

var boxPlot = myPlot.Add.Box(boxes);

boxPlot.Groups.Orientation = Orientation.Horizontal;

myPlot.SavePng("horizontal-box-plots.png");

Box Plot Series

Similarly to bar charts, box plots can be compared across multiple categories.

ScottPlot.Plot myPlot = new();

Random rand = new(0);

ScottPlot.Plottables.Box CreateBox()
{
    int N = 50;
    double mean = rand.NextDouble() * 3;
    double stdDev = rand.NextDouble() * 3;

    double[] values = Generate.RandomNormal(N, mean, stdDev);
    Array.Sort(values);
    double min = values[0];
    double q1 = values[N / 4];
    double median = values[N / 2];
    double q3 = values[3 * N / 4];
    double max = values[N - 1];

    return new ScottPlot.Plottables.Box
    {
        WhiskerMin = min,
        BoxMin = q1,
        BoxMiddle = median,
        BoxMax = q3,
        WhiskerMax = max,
    };
}

int numBoxesPerSeries = 3;
int numSeries = 2;
ScottPlot.Plottables.BoxGroup[] series = new ScottPlot.Plottables.BoxGroup[numSeries];
var colorPalette = new Category10();
for (int i = 0; i < series.Length; i++)
{
    series[i] = new ScottPlot.Plottables.BoxGroup
    {
        Fill = new FillStyle
        {
            Color = colorPalette.Colors[i % colorPalette.Colors.Length]
        },
        Boxes = new ScottPlot.Plottables.Box[numBoxesPerSeries],
    };

    for (int j = 0; j < series[i].Boxes.Count; j++)
    {
        var box = CreateBox();
        box.Position = j;
        series[i].Boxes[j] = box;
    }
}

var boxPlot = myPlot.Add.Box(series);

myPlot.SavePng("box-plot-series.png");