Crashing QwtPlot

Hi Everyone,

I am working on qwtplot, and its getting crash, as I'm trying to overload with data.
I have read on this link http://www.qtforum.de/viewtopic.php?t=16041
And come to the conclusion that I should use setRawSamples() rather than using setSamples().

So previously I'm doing this,

Code:

---

QVector<QPointF> m_points;
//Adding some data to m_points
m_points.push_back(QPointF(xValue, yValue));
setSamples(m_points);

---

And now I'm trying with this below, but getting some error

Code:

---

QVector<QPointF> m_points;
//Adding some data to m_points
m_points.push_back(QPointF(xValue, yValue));
setRawSamples(m_points.at(0).x(), m_points.at(0).y(), sizeof(m_points));

---

Quote:

---

Error:
error: C2664: 'void QwtPlotCurve::setRawSamples(const double *,const double *,int)': cannot convert argument 1 from 'qreal' to 'const double *'

---

In the above code I tried with casting but doesn't working.

First of all I want to try this and want to check does is this the problem that crashing my application.
If someone think, may be their is other problem then please mentioned that also.

Thanks.

setRawSamples can avoid creating an extra copy of the data, but then you have to take care of the lifetime of your buffers.

So the posted code for the second implementation is totally wrong:

- it does not even compile ( passing doubles, where double* is expected )
- the vector is on the stack ( = not caring of the lifetime )
- the vector stores its values [x][y][x][y].. so there are no [x][x].. [y][y].. arrays inside

Also note, that Qt containers do copy on write. So using setSamples already avoids copying, what makes the second implementation pointless.

Uwe

You right, the second implementation is not even compile and not even written right. The thing is that I'm unable to right it proper way, that's the other thing.
And I came to this conclusion because, the line setRawSamples() is working fine if the data is not too fast, but if I given bulk or fast data its crashing at this point.

Code:

---

void Input::plottingData(QVariant data)
{
    if (data.userType() == QMetaType::QPointF)
    {
        m_points.push_back(data.toPointF());
    }
    if (data.userType() == QMetaType::Double)
    {
        double yValue = data.toDouble();
 
        cframe::TimeStamp time = Module::createTimestamp();
        qint64 nanoElapsed = time.nSecsElapsed();
        double xValue = nanoElapsed/(double)1000000000;
        m_points.push_back(QPointF(xValue, yValue));
    }
    if (data.userType() == qMetaTypeId<QVector<QPointF>>())
    {
        QVector<QPointF> points = data.value<QVector<QPointF>>();
        m_points.append(points);        
    }
 
    m_curve.setSamples(m_points);
}

---

Can you please check this code above and clarify does the problem is in the coding somewhere ?

Thanks

Their is typo mistake,

Quote:

---

Originally Posted by npatil15

And I came to this conclusion because, the line setRawSamples() is working fine if the data is not too fast, but if I given bulk or fast data its crashing at this point.

---

In the above statement, its setSamples() not setRawSamples()

The code is o.k.

Okay, thanks.
The I think the crashing issue is not here then,
Still I need to debug my rest of code, why my plot is crashing during fast and bulk amount of data.

Is their anything you can suggest to look over ?

Ask your debugger, use the sanitizer ...