Input file of 14 points

Results in QGIS

Results in geojson format

Patch

Hi, first of all thanks a lot for your interest in the Voronoi service and for providing such kinf of feedbacks.

I have to conceede that I do not take a look at the CGAL service for long time now. Nevertheless, by searching on the internet I realized that we are nor using any constraint on our computation (based on the data extent) nor treating the rays which may occurs when producing the Voronoi diagram through the CGAL library.

I would like to invite you to apply the patch linked in this page and let me know if it solves your issue with your dataset.

Note that I got 5 rays using your dataset and I treat them in a really stupid manner by now, but this is just for confirming that the rays are well the missing "edges" you were talking about in your original post.

I would like to invite you to let us know if this solves your issue, in this case we may think of trying to provide better solution or use contrained computation. Please take a look at the ​CGAL Triangulation Documentation

I forgot to mention that I've setup a brand new ZOO-Kernel on the ZOO-Project server with the patched version of the service and it is available from here :

http://svn.zoo-project.org/cgi-bin/cgal/zoo_loader.cgi

I've also copied your GML data in dl directory so this request should work :

http://svn.zoo-project.org/cgi-bin/cgal/zoo_loader.cgi?request=Execute&service=WPS&version=1.0.0&Identifier=Voronoi&DataInputs=InputPoints=Reference@xlink:href=http://www.zoo-project.org/dl/london.gml&RawDataOutput=Result

Showing the Rays as well

Yes! the missing edge is one of the rays - I've attached a picture with them all showing.