Hi Lie Pablo, please see the image at https://dl.dropboxusercontent.com/u/32772/pythonOCC/findverticesnearedge.png
I have a set of edges (blue) that intersect the surface at an angle close to but not exactly 90° and the coordinates of the mesh nodes. What I'm trying to do is assign nodes in the vicinity of an edge to that edge. Nodes further away than a certain tolerance value should not be assigned at all. The BRepExtrema_DistShapeShape approach is working fine, but it is very slow and I'm looking for faster alternatives. The part in my code with the sqrt calculations is only because the nodes need to be moved parallel to the assigned edge and i need a unit vector for that. It doesn't slow things down too much. Uwe On 06/08/13 18:57, Lie Pablo Grala Pinto wrote: > I have done things like you are trying to do for very big meshes. But I > didn't understand correctly what is your problem. > > Does your problem is finding point (nodes) to curve distance? Or edge to > edge distance? > > When you compare distances, a way to compute faster is to scape from > sqrt. For example > > Slow way to compare the distance of a_point to a point-set: > For each point in point_set; sqrt(dot(point-a_point,point-a_point) < > distance; > > Faster way to compare the distance of a_point to a point-set: > For each point in point_set; dot(point-a_point,point-a_point) < > distance**2; > > Even faster way to compare the distance of a_point to a point-set: > temp=distance**2; For each point in point_set; > dot(point-a_point,point-a_point) < temp; > > Good luck > Lie Pablo > > > > On 6 August 2013 13:21, Uwe Schlifkowitz <schlifkow...@intales.com > <mailto:schlifkow...@intales.com>> wrote: > > Hi all, > > I am trying to find vertices that are near an edge within some kind of > tolerance. > > Imagine lots of short edges intersecting a surface. The edges are > parallel to the surface normal and evenly distributed over the surface. > The surface has a mesh and I have a dictionary of said edges and a > dictionary of mesh nodes (vertices). > > So what I'm trying to do is assign all nodes that are within ~3mm of an > edge to that edge. > > I used BRepExtrema_DistShapeShape which seems to do the job quite well > for small numbers of points and edges (600 points, 72 edges). > For larger numbers however, the process consumes all available memory > (16GB) and then hangs the machine. > > So my questions are: > * Is there a better way to achieve what i want? > * How can I reduce memory consumption? > > Unrelated question: why does the conversion from TopoDS_Vertex to gp_Pnt > fail when i write TopoDS_Vertex instead of TopoDS_vertex? What exactly > is the difference? If it's in the docs I must have missed it. > > > Thanks, > Uwe > > #--------------------------------------- > pntDic = {} # some dictionary of gp_Pnt > curveDic = {} # some dictionary of TopoDS_Edge > tol = 3 > curveToNode = {} > for k in pntDic: > for l in curveDic: > v = BRepBuilderAPI_MakeVertex(pntDic[k]) > dist = BRepExtrema_DistShapeShape(v.Vertex(), curveDic[l]) > if dist.Value() <= tol: > # assign curve to node > # get normalized edge direction > # get start and end vertex of edge > ex = TopExp_Explorer(curveDic[l], TopAbs_VERTEX) > > # no need for a loop since we know for a fact that > # the edge has only one start and one end > c = ex.Current() > cv = TopoDS_vertex(c) > v0 = BRep_Tool_Pnt(cv) > ex.Next() > c = ex.Current() > cv = TopoDS_vertex(c) > v1 = BRep_Tool_Pnt(cv) > dx = v1.X() - v0.X() > dy = v1.Y() - v0.Y() > dz = v1.Z() - v0.Z() > edgeScaling = sqrt(dx ** 2 + dy ** 2 + dz ** 2) > edgeDirection = [dx / edgeScaling, > dy / edgeScaling, > dz / edgeScaling] > curveToNode[k] = (l, edgeDirection) > del ex, v1, v0, dx, dy, dz, c, cv > del edgeScaling, edgeDirection > else: > pass > del v, dist > #--------------------------------------- > > > -- > INTALES GmbH Engineering Solutions > Uwe Schlifkowitz -- Software Engineer > Innsbrucker Str. 1, A-6161 Natters > Phone: +43 512 54611120 <tel:%2B43%20512%2054611120> -- > schlifkow...@intales.com <mailto:schlifkow...@intales.com> > > _______________________________________________ > Pythonocc-users mailing list > Pythonocc-users@gna.org <mailto:Pythonocc-users@gna.org> > https://mail.gna.org/listinfo/pythonocc-users > > > > > _______________________________________________ > Pythonocc-users mailing list > Pythonocc-users@gna.org > https://mail.gna.org/listinfo/pythonocc-users > _______________________________________________ Pythonocc-users mailing list Pythonocc-users@gna.org https://mail.gna.org/listinfo/pythonocc-users
