如何在条形图上使用单独的范围轴绘制两个数据系列

Question

I was wondering if it is possible to plot ( XYPlot ) a dataset in JFreeChart where the series contained in the dataset are plotted against separate y-axis ranges. The date is clustered by category/timestamp, eg if my dataset was this:

Timestamp        Val1        Val2
2019-04-26       0.6         603
2019-04-25       2.1         1040
2019-04-24       4.1         255

It is impractical to plot both value series on the same range axis.

I've attempted extracting each series into its own dataset, so that I can call plot.mapDataSetToRangeAxis() ; but when I add multiple datasets to the plot, the bars tend to render on top of each other. Perhaps I'm missing something simple?

There are a few posts that address separate elements of what I'm looking for, but think I need something that combines these two:

• JFreeChart - XYBarChart Show Separate Bars for Each Series
• Setting different y-axis for two series with JFreeChart

Here is the python code I'm currently using—inside inductive automation/ignition's reporting module; they allow you configure the JFreeChart prior to rendering.

def configureChart(chart):
    from org.jfree.chart.axis import DateAxis
    from org.jfree.data.xy import XYSeries, XYSeriesCollection
    from org.jfree.chart.renderer.xy import ClusteredXYBarRenderer
    from java.awt import Color
    from java.text import NumberFormat

    class mins_to_str(NumberFormat):
        def format(self,*args,**kwargs):
            r = ''
            number = args[0]
            hrs = number//60
            mins = number%60
            r = '%02i:%02i' %(hrs,mins)
            if len(args)>1:
                toAppendTo = args[1]
                pos = args[2].getField()
                r = toAppendTo.insert(pos,r)
            return r

    plt = chart.getPlot()

    renderer = ClusteredXYBarRenderer
    xax = DateAxis()
    plt.setDomainAxis(xax)

    for i in range(plt.getDatasetCount()):
        d = plt.getDataset(i)
        dsc = XYSeriesCollection()
        series = XYSeries(d.getSeriesKey(0))
        print('SERIES [%s]' %series)
        for r in range(d.getItemCount(0)):
            xv = d.getXValue(0,r)
            yv = d.getYValue(0,r)
            print('  X: %s (%s)' %(xv,type(xv)))
            print('  Y: %s (%s)' %(yv,type(yv)))
            series.add(xv,yv)
        dsc.addSeries(series)
        plt.setDataset(i,dsc) # assuming all of my series need to be in the same dsc for this to work...
        plt.setRenderer(i,renderer)

        if i > 0:
            plt.mapDatasetToRangeAxis(i,1)
        else:
            plt.mapDatasetToRangeAxis(i,0)  

    plt.getRangeAxis(0).setNumberFormatOverride(mins_to_str())

Currently, I'm getting this:

Any ideas/help would be greatly appreciated.

Answer 1

Here is a time series chart within ignition's Report developing environment.

This one is completely setup using the Chart Options and just a tiny bit of JFreeChart scripting. Should be similar to designing a bar chart. Hope it helps.

JFreeChart Script -

    #Import Java classes - Color & JFreeChart
from java.awt import Color
from org.jfree.chart.plot import CategoryPlot

#get Plot of Current Chart
plot = chart.getPlot()

#Set color of domain and range gridlines
plot.setDomainGridlinePaint(Color.black)
plot.setRangeGridlinePaint(Color.black)