我有一个子计算图形的数组,另一个子图形图。 有两个线程用于计算(getArrayForGraph()),另一个用于绘制图形。
现在我希望图形沿着侧面阵列更新功能绘制,因为阵列非常长(10 000 000),使整个过程非常慢。
如果图表可用点,我将如何开始更新图表。或哪种解决性能问题的最佳方法。
@Test
public void test(){
Map<String, Integer> a = new HashMap<>();
a.put("x", new Integer(0)); // Store Integer(0)
Integer i = a.get("x"); // Get it
a.put("x", i++); // Unbox it, rebox it, store it again (still 0); increment it afterward
i = a.get("x"); // Same again
a.put("x", i++); // ...
i = a.get("x"); // ...
a.put("x", i++); // Unbox it, rebox it, store it again (still 0); increment it afterward
System.err.println(i); // Show it; contains the *incremented* value
}
答案 0 :(得分:0)
这是一个完整的例子。它使用BackgroundWorker来执行工作,该工作向UI发送信号以进行更新。它还汇总了数据。我选择只显示2000点因为1. MSChart无法处理很多点,而且2.我的显示器只有1920像素宽,因此没有必要尝试显示更多。此示例将为每5k(10M / 2k)点绘制1个点,即x:最新时间,y:平均值。如果你愿意,你也可以采取平均时间,但这个情节更整洁。
它不包含您的特定变量,因为缺少某些信息。但是你应该知道它是如何工作的,并使它在你的环境中工作
Public Class Form1
Private worker As System.ComponentModel.BackgroundWorker
Private chartX As New List(Of Long)()
Private chartY As New List(Of Long)()
' settings for demo
Private count As Long = 10000000
Private numberOfGroups As Long = 2000
Private groupSize As Long = count / numberOfGroups
Private Sub doWork(sender As Object, e As System.ComponentModel.DoWorkEventArgs)
' holds chunk of data to be grouped
Dim temp As New List(Of Tuple(Of Long, Long))()
For i As Long = 1 To count
' generate some scientific data!
temp.Add(New Tuple(Of Long, Long)(i, Convert.ToInt64(i * Math.Sin(Math.PI * i / count) * Math.Sin(i))))
' make this happen 10 times
If i Mod count / 10 = 0 Then
' group the time and data into same groupings
Dim aggregate =
temp.
Select(Function(d, index) New Tuple(Of Integer, Tuple(Of Long, Long))(index, d)).
GroupBy(Function(d) Math.Floor(d.Item1 / groupSize))
' take the latest time
Dim xAggregate =
aggregate.
Select(Function(a) a.Max(Function(b) b.Item2.Item1)).
ToArray()
' take the average data
Dim yAggregate =
aggregate.
Select(Function(a) Convert.ToInt64(a.Average(Function(b) b.Item2.Item2))).
ToArray()
' clear temp data
temp.Clear()
' -- calling resetChartData() here would make the chart display only the currently processed data
' -- might be useful if you were displaying data from an oscilloscope for instance
' resetChartData()
' add the aggregate data to the chart source data
chartX.AddRange(xAggregate)
chartY.AddRange(yAggregate)
' report progress (initiate charting)
DirectCast(sender, System.ComponentModel.BackgroundWorker).ReportProgress(Convert.ToInt32(100 * i / count))
End If
Next
End Sub
Private Sub progressChanged(sender As Object, e As System.ComponentModel.ProgressChangedEventArgs)
' ui thread runs here.
' perform minimal ui operations as not to slow down ui thread
' added a progress bar, nice to see
ProgressBar1.Value = e.ProgressPercentage
' databind the data
Chart1.Series("Series1").Points.DataBindXY(chartX, chartY)
End Sub
Private Sub resetChartData()
chartX.Clear()
chartY.Clear()
End Sub
Private Sub runWorkerCompleted(sender As Object, e As System.ComponentModel.RunWorkerCompletedEventArgs)
Button1.Enabled = True
End Sub
Private Sub Button1_Click(sender As Object, e As EventArgs) Handles Button1.Click
' to prevent double clicks
Button1.Enabled = False
worker = New System.ComponentModel.BackgroundWorker()
AddHandler worker.DoWork, AddressOf doWork
AddHandler worker.ProgressChanged, AddressOf progressChanged
AddHandler worker.RunWorkerCompleted, AddressOf runWorkerCompleted
worker.WorkerReportsProgress = True
' do work on background thread
worker.RunWorkerAsync()
End Sub
Private Sub Form1_Load(sender As Object, e As EventArgs) Handles MyBase.Load
' set up chart
Chart1.Series.Add("Series1")
Chart1.Series.Single().MarkerStyle = DataVisualization.Charting.MarkerStyle.None
Chart1.Series.Single().ChartType = DataVisualization.Charting.SeriesChartType.FastLine
Chart1.Legends.Clear()
End Sub
End Class