QCustomPlot技术操作：已有绘图加散点及按X值取Y值绘图

Hey there! Let's break down your two QCustomPlot questions with practical code examples, building off your existing createPlot() function.

To add scatter points above your already-created plot, you'll need to create a dedicated QCPGraph (or repurpose an existing one) to hold the scatter data, define a distinct scatter style, and then refresh the plot. Here's how to implement it:

Example Code #

void Qt_PlotTest::addScatterPoints(QCustomPlot* customPlot) {
    // Create a new graph to hold scatter points (keeps your original plot intact)
    QCPGraph* scatterGraph = customPlot->addGraph();
    
    // Define your scatter style: red circles, size 8
    QCPScatterStyle scatterStyle(QCPScatterStyle::ssCircle);
    scatterStyle.setColor(Qt::red);
    scatterStyle.setSize(8);
    scatterGraph->setScatterStyle(scatterStyle);
    
    // Hide the line for this graph (we only want to see scatter points)
    scatterGraph->setLineStyle(QCPGraph::lsNone);
    
    // Prepare your scatter data (adjust X/Y values to match your use case)
    double now = QDateTime::currentDateTime().toTime_t();
    QVector<double> scatterX, scatterY;
    scatterX << now - 7200 << now - 3600 << now; // 2 hours ago, 1 hour ago, now
    scatterY << 45 << 60 << 70; // Corresponding Y values for your scatter points
    
    // Assign data and refresh the plot
    scatterGraph->setData(scatterX, scatterY);
    customPlot->rescaleAxes(true); // Optional: auto-adjust axes to fit new points
    customPlot->replot();
}

How to Use It #

Call this function after your createPlot() function finishes initializing the plot, or trigger it from a user action (like a button click):

// Inside your createPlot() function, after setting up the main plot
addScatterPoints(customPlot);

If you want to pull Y values corresponding to specific X values from your existing plot data, you can use QCPGraph's built-in search methods, or interpolate values if your target X isn't an exact match in the original data. Here's a step-by-step implementation:

Example Code #

void Qt_PlotTest::plotPointsFromExistingX(QCustomPlot* customPlot) {
    // Get your main plot graph (assuming it's the first graph at index 0)
    QCPGraph* mainGraph = customPlot->graph(0);
    if (!mainGraph) return; // Exit if no main plot exists
    
    // Define the X values you want to look up
    double now = QDateTime::currentDateTime().toTime_t();
    QVector<double> targetXValues = {now - 7200, now - 3600, now};
    
    QVector<double> plotX, plotY;
    
    for (double targetX : targetXValues) {
        // Find the first data point with an X value >= targetX
        int dataIndex = mainGraph->findBegin(targetX, true);
        
        if (dataIndex >= 0 && dataIndex < mainGraph->data()->size()) {
            // Case 1: Exact X value match
            if (qFuzzyCompare(mainGraph->data()->at(dataIndex)->key, targetX)) {
                plotX << targetX;
                plotY << mainGraph->data()->at(dataIndex)->value;
            }
            // Case 2: Interpolate Y value if X is between two existing points
            else if (dataIndex > 0) {
                double x1 = mainGraph->data()->at(dataIndex - 1)->key;
                double y1 = mainGraph->data()->at(dataIndex - 1)->value;
                double x2 = mainGraph->data()->at(dataIndex)->key;
                double y2 = mainGraph->data()->at(dataIndex)->value;
                
                // Linear interpolation to get approximate Y value
                double interpolatedY = y1 + (y2 - y1) * (targetX - x1) / (x2 - x1);
                plotX << targetX;
                plotY << interpolatedY;
            }
        }
    }
    
    // Create a graph to draw these points
    QCPGraph* lookupPointsGraph = customPlot->addGraph();
    QCPScatterStyle pointStyle(QCPScatterStyle::ssSquare);
    pointStyle.setColor(Qt::darkBlue);
    pointStyle.setSize(10);
    lookupPointsGraph->setScatterStyle(pointStyle);
    lookupPointsGraph->setLineStyle(QCPGraph::lsNone);
    
    // Assign data and refresh
    lookupPointsGraph->setData(plotX, plotY);
    customPlot->replot();
}

Key Notes #

• qFuzzyCompare is used to compare floating-point X values safely (avoids precision issues).
• Linear interpolation is useful if your target X doesn't exactly match any point in the original dataset.
• If you don't want to create a new graph every time, you can initialize a single scatter graph once and update its data with setData() later.

内容的提问来源于stack exchange，提问作者remi