The plugin causes a segfault

[pcav]

Both when writing to memory layer and to a shapefile.
It does not happen on all PCs on a course here, but we had it both on Win and on Debian.

[jdugge]

Could you provide a shapefile and the parameters used to trigger the segfault?

[jdugge]

Maybe this is caused by the iterative method not converging. Could you try replacing the secant method in bufferbypercentage.py with the following:

def secant(func, oldx, x, *args, **kwargs):
    """Find the root of a function"""
    TOL = kwargs.pop('TOL', 1e-6)
    oldf, f = func(oldx, *args), func(x, *args)
    iterations = 0
    if (abs(f) > abs(oldf)):
        oldx, x = x, oldx
        oldf, f = f, oldf
    while iterations < 100:
        dx = f * (x - oldx) / float(f - oldf)
        if abs(dx) < TOL * (1 + abs(x)):
            return x - dx
        oldx, x = x, x - dx
        oldf, f = f, func(x, *args)
        iterations += 1
    print "Maximum number of iterations reached, no convergence."
    return x

[pcav]

sorry, no zip upload here on GH. I can send it by email if you wish.

[pcav]

Sorry, I'm on a course now, very limited time to do anything. We'll be happy of testing a new version though.
Thanks.

[jdugge]

I've created a new branch that catches if there's an endless loop and tells the user which feature caused the problem. Could you download the directory as a .zip and put that in the ~/.qgis2/.python/plugins directory (you'll have to rename the folder from BufferByPercentage-limit_iterations to BufferByPercentage) and see if the segfault goes away?

Thanks for your help in this!

[pcav]

Sorry, 0 time available in this period - just overcommitted.
Thanks for looking into this.