I want to create a bar-chart, but extraordinary high values should be shortened. An example is this image:
(source: epa.gov)
I hope it is clear what I want.
My question is: How can I do this with JFreeChart. If it isn't possible with JFreeChart you can possibly recommend alternative open-source Java-libraries to produce such an output.
You could do it with a CombinedDomainCategoryPlot
or CombinedDomainXYPlot
. Set the range axis of the first plot to your cut off value and then do something similar with the second plot. Then add them to a combined plot.
import org.jfree.chart.ChartFactory;
import org.jfree.chart.ChartFrame;
import org.jfree.chart.JFreeChart;
import org.jfree.chart.axis.NumberAxis;
import org.jfree.chart.plot.CategoryPlot;
import org.jfree.chart.plot.CombinedDomainCategoryPlot;
import org.jfree.chart.plot.PlotOrientation;
import org.jfree.data.category.DefaultCategoryDataset;
public class PlayChart {
public static void main(String[] args) {
DefaultCategoryDataset ds = new DefaultCategoryDataset();
ds.addValue(100, "A", "A");
ds.addValue(200, "A", "B");
ds.addValue(400, "A", "C");
ds.addValue(500, "A", "D");
ds.addValue(2000, "A", "E");
JFreeChart bc = ChartFactory.createBarChart("My Bar Chart", "Things", "Counts", ds, PlotOrientation.VERTICAL, true, false, false);
JFreeChart bcTop = ChartFactory.createBarChart("My Bar Chart", "Things", "Counts", ds, PlotOrientation.VERTICAL, true, false, false);
CombinedDomainCategoryPlot combinedPlot = new CombinedDomainCategoryPlot();
CategoryPlot topPlot = bcTop.getCategoryPlot();
NumberAxis topAxis = (NumberAxis) topPlot.getRangeAxis();
topAxis.setLowerBound(1500);
topAxis.setUpperBound(2000);
combinedPlot.add(topPlot, 1);
CategoryPlot mainPlot = bc.getCategoryPlot();
combinedPlot.add(mainPlot, 5);
NumberAxis mainAxis = (NumberAxis) mainPlot.getRangeAxis();;
mainAxis.setLowerBound(0);
mainAxis.setUpperBound(600);
JFreeChart combinedChart = new JFreeChart("Test", combinedPlot);
ChartFrame cf = new ChartFrame("Test", combinedChart);
cf.setSize(800, 600);
cf.setVisible(true);
}
}
The plots will share the same X-axis. You'll need to play with the renderers to set colours and labels.
removed dead ImageShack link