Box Plot Charts
Box Plot Charts are useful for displaying variations in statistical samples of data and data details in a small space.
Getting Started
The Box Plot Chart uses the following values:
- First and third quartile (q1 and q3)
- Median (that is the second quartile)
- Upper and lower value (inner fences)
- Mean and outliers
To create a Box Plot series, use boxPlot
in the series
configuration.
Binding the Box Plot Chart
When you bind the Box Plot Chart, you have to set the following special properties of the Series
object so that the Chart is successfully displayed:
- (Mandatory)
q1Field
(first quartile) - (Mandatory)
q3Field
(third quartile) - (Mandatory)
lowerField
- (Mandatory)
upperField
- (Mandatory)
medianField
(second quartile) - (Optional)
meanField
- (Optional)
outliersField
All properties except for outliersField
have to point to numeric fields in the DataSource. outliersField
has to point to a field that contains an array of numbers.
<div id="chart"></div>
<script>
$("#chart").kendoChart({
series: [{
type: 'boxPlot',
q1Field: 'q1',
q3Field: 'q3',
lowerField: 'lower',
upperField: 'upper',
medianField: 'median',
meanField: 'mean',
outliersField: 'outliers'
}],
dataSource: {
data: [
{ year: 2004, lower: 1.8, q1: 2.75, median: 3.35, q3: 3.825, upper: 4.9, mean: 3.4 },
{ year: 2005, lower: 1.7, q1: 2.275, median: 3.2, q3: 3.825, upper: 5.5, outliers: [0.5, 6.7] },
{ year: 2006, lower: 1.2, q1: 1.95, median: 2.45, q3: 3.075, upper: 3.5, mean: 2.5 },
{ year: 2007, lower: 1.3, q1: 1.9, median: 3.05, q3: 3.425, upper: 4, mean: 2.7, outliers: [7, 8.5] }
]
},
categoryAxis: {
field: 'year'
},
tooltip: {
visible: true
}
});
</script>
Setting the Orientation
You can change the orientation of the Box Plot Chart by setting the series type to verticalBoxPlot
.
<div id="chart"></div>
<script>
$("#chart").kendoChart({
series: [{
type: 'verticalBoxPlot',
q1Field: 'q1',
q3Field: 'q3',
lowerField: 'lower',
upperField: 'upper',
medianField: 'median',
meanField: 'mean',
outliersField: 'outliers'
}],
dataSource: {
data: [
{ year: 2004, lower: 1.8, q1: 2.75, median: 3.35, q3: 3.825, upper: 4.9, mean: 3.4 },
{ year: 2005, lower: 1.7, q1: 2.275, median: 3.2, q3: 3.825, upper: 5.5, outliers: [0.5, 6.7] },
{ year: 2006, lower: 1.2, q1: 1.95, median: 2.45, q3: 3.075, upper: 3.5, mean: 2.5 },
{ year: 2007, lower: 1.3, q1: 1.9, median: 3.05, q3: 3.425, upper: 4, mean: 2.7, outliers: [7, 8.5] }
]
},
categoryAxis: {
field: 'year'
},
tooltip: {
visible: true
}
});
</script>
Customizing the Outliers
Outliers are values that appear outside the range between lower and upper values (inner fences). Based on their distance from the inner fences, the outliers in a box plot are mild and extreme. The Box Plot Chart renders them by default as crosses (mild) and circles (extreme).
You can configure the outliers separately and customize them by using:
-
series.outliers
for mild outliers. -
series.extremes
for extreme outliers.
<div id="chart"></div>
<script>
$("#chart").kendoChart({
series: [{
type: 'verticalBoxPlot',
q1Field: 'q1',
q3Field: 'q3',
lowerField: 'lower',
upperField: 'upper',
medianField: 'median',
meanField: 'mean',
outliersField: 'outliers',
outliers: {
border: {
color: "green"
},
type: "triangle"
},
extremes: {
border: {
color: "red"
},
type: "square",
rotation: 45
}
}],
dataSource: {
data: [
{ year: 2004, lower: 1.8, q1: 2.75, median: 3.35, q3: 3.825, upper: 4.9, mean: 3.4 },
{ year: 2005, lower: 1.7, q1: 2.275, median: 3.2, q3: 3.825, upper: 5.5, outliers: [0.5, 6.7] },
{ year: 2006, lower: 1.2, q1: 1.95, median: 2.45, q3: 3.075, upper: 3.5, mean: 2.5 },
{ year: 2007, lower: 1.3, q1: 1.9, median: 3.05, q3: 3.425, upper: 4, mean: 2.7, outliers: [7, 8.5] }
]
},
categoryAxis: {
field: 'year'
},
tooltip: {
visible: true
}
});
</script>