JavaFX LineChart具有65000个数据点时的性能问题
它需要JavaFX 15分钟来构建描述的LineChart，它不适用于我的任务。
使用良好的旧Swing和jFreeChart的类似实现需要1.5秒来构建图表。
但是我仍然想要实现一个JavaFX。
这是我的代码：
public class FXMLController implements Initializable {
@FXML
private Label statusbar;
@FXML
public LineChart lineChart;
@FXML
public Button connect;
@FXML
public MenuItem options;
@FXML
public NumberAxis xAxis;
@FXML
NumberAxis yAxis;
@FXML
private void connect(ActionEvent event) {
}
public static FileChooser fileChooser = new FileChooser();
public static String path;
public static XYChart.Series<Integer, Integer> dataSeries = new XYChart.Series<Integer, Integer>();
public static int y = 0;
public static XYChart.Data<Integer, Integer> data;
@FXML
private void open(ActionEvent event) {
fileChooser.setTitle("Open Resource File");
fileChooser.getExtensionFilters().addAll(
new ExtensionFilter("Text Files", "*.txt"),
new ExtensionFilter("Image Files", "*.png", "*.jpg", "*.gif"),
new ExtensionFilter("Audio Files", "*.wav", "*.mp3", "*.aac"),
new ExtensionFilter("All Files", "*.*"));
File selectedFile = fileChooser.showOpenDialog(new Stage());
if (selectedFile != null) {
path = selectedFile.getAbsolutePath();
System.out.println(path);
try {
ReadingFromFile.readFile(path);
} catch (IOException ex) {
Logger.getLogger(FXMLController.class.getName()).log(Level.SEVERE, null, ex);
}
}
}
@FXML
private void close(ActionEvent event) {
}
@FXML
private void getconnect(ActionEvent event) {
}
@Override
public void initialize(URL url, ResourceBundle rb) {
xAxis.setLabel("Tick");
xAxis.setTickUnit(100);
yAxis.setLabel("Signal");
xAxis.setForceZeroInRange(false);
lineChart.setLegendVisible(false);
lineChart.setCreateSymbols(false);
lineChart.setAnimated(false);
lineChart.getData().add(dataSeries);
}
}
并从文件中读取：
public class ReadingFromFile extends FXMLController {
public static String s = null;
public static String[] str;
public static int parseInt;
public static void readFile(String filename)
throws IOException {
BufferedReader br = new BufferedReader(new FileReader(filename));
try {
StringBuilder sb = new StringBuilder();
String line = br.readLine();
while (line != null) {
sb.append(line);
sb.append(System.lineSeparator());
line = br.readLine();
System.out.println(line);
try {
str = line.split(" ");
for (int i = 0; i < str.length; i = i + 2) {
s = str[i + 1] + str[i];
parseInt = Integer.parseInt(s, 16);
javafx.application.Platform.runLater(new Runnable() {
@Override
public void run() {
data = new XYChart.Data<Integer, Integer>(y, parseInt);
//data.setNode(new HoveredThresholdNode(0, second, ""));
dataSeries.getData().add(data);
y++;
}
});
}
} catch (java.lang.NullPointerException ex) {
System.out.println("тут ноль!!!");
}
}
} finally {
br.close();
}
}
}
我遇到过类似的问题，每隔几秒就为LineChart添加100,000点。 我们使用Ramer-Douglas-Peucker算法解决了这个问题，这可以减少线路中的点数，而无需用户注意。 我在LGPL许可下的JTS拓扑套件中找到了现成的实现。
这是我的测试代码。
public class ChartUpdate extends Application {
public static void main(String[] args) {
launch(args);
}
@Override
public void start(Stage stage) {
NumberAxis xAxis = new NumberAxis(0, 50_000, 5000);
xAxis.setAutoRanging(false);
NumberAxis yAxis = new NumberAxis(-1, 1, 25);
yAxis.setAutoRanging(false);
LineChart<Number, Number> graph = new LineChart<>(xAxis, yAxis);
graph.setAnimated(false);
graph.setCreateSymbols(false);
graph.setLegendVisible(false);
Series<Number, Number> series = new Series<>();
stage.setScene(new Scene(graph));
GeometryFactory gf = new GeometryFactory();
long t0 = System.nanoTime();
Coordinate[] coordinates = new Coordinate[100_000];
for (int i = 0; i < coordinates.length; i++) {
coordinates[i] = new Coordinate(i, Math.sin(Math.toRadians(i / 100)));
}
Geometry geom = new LineString(new CoordinateArraySequence(coordinates), gf);
Geometry simplified = DouglasPeuckerSimplifier.simplify(geom, 0.00001);
List<Data<Number, Number>> update = new ArrayList<Data<Number, Number>>();
for (Coordinate each : simplified.getCoordinates()) {
update.add(new Data<>(each.x, each.y));
}
long t1 = System.nanoTime();
System.out.println(String.format("Reduces points from %d to %d in %.1f ms", coordinates.length, update.size(),
(t1 - t0) / 1e6));
ObservableList<Data<Number, Number>> list = FXCollections.observableArrayList(update);
series.setData(list);
graph.getData().add(series);
stage.show();
}
}
Ramer-Douglas-Peucker不必要的复杂，即使采用更快的下采样策略，仅靠这一点也不足以获得我们所需的性能。 在这里查看我的答案以获得更完整的解决方案 这为我实现了40,000左右的实时更新。
链接地址: http://www.djcxy.com/p/25163.html上一篇: Performance issue with JavaFX LineChart with 65000 data points