[Bug] VolumeEnclosed give NaNs when it has a point coincident with its origin

[ConduitDan]

import meshtools
import plot
import optimize

var vertexlist = [[0.5,0.5,0.5], [0.5,1.5,0.5], [1.5,1.5,0.5], [1.5,0.5,0.5], [0.5,0.5,1.5],[0.5,1.5,1.5], [1.5,1.5,1.5], [1.5,0.5,1.5]]

var facelist = [[0,3,2,1], [0,4,5,1], [0,3,7,4], [2,3,7,6], [4,5,6,7], [1,2,6,5]]

var m = PolyhedronMesh(vertexlist, facelist)

//Show(plotmesh(m, grade=[0,1,2]))

print(m)

var problem = OptimizationProblem(m)

var la = Area()
var lv = VolumeEnclosed()

problem.addenergy(la)
problem.addconstraint(lv)

var opt = ShapeOptimizer(problem, m)

var mr=MeshRefiner([m])
var refmap = mr.refine()
for (el in [problem, opt]) el.update(refmap)
m = refmap[m]

opt.relax(100)

Show(plotmesh(m, grade=[0,1,2]))

[joshichaitanya3]

Hmm, interesting. The Polyhedronmesh looks fine.

I noticed that
print lv.total(m)
returns 2 in this case, while it should return 1.

The cube in the cube.morpho file returns 1 correctly.

[softmattertheory]

@ConduitDan move all the vertices so that (0,0,0) lies in the cube. Does it still fail then?

The problem I think is that VolumeEnclosed isn't guaranteed to work if (0,0,0) isn't in the volume (until we've done Mesh orientation)

[ConduitDan]

VolumeEnclosed should probably throw and error if it encounters a point coincident with the origin that it is measuring volume from

[ConduitDan]

Alternatively we should protect from divide by zero by adding eps to the denominator

[softmattertheory]

This has now been fixed in v0.6.2. Thanks so much for raising this @ConduitDan!