Hello all,
Does anybody know any framework/code for IGD indicator, GD indicator in python or julia programming?.Also Is it ok to use a hypervolume indicator when it is biased with respect to reference points?
This may help in IGD calculation
def calculateigd(truepf, pf): """Calculate IGD value for truepf and pf Parameters ---------- truepf: "true" pf value pf: estimated pf value Returns ------- igd: IGD value """ Y = scipy.spatial.distance.cdist(truepf, pf, 'euclidean') mindist = np.min(Y, axis=1) igd = np.mean(mindist) return igd
Here is another implementation
please see the attached file for both of them
def IGD(truePF, PF):
distEuclidean = []
sumMin = 0
for i in truePF:
for x in PF:
d = distance.euclidean(i, x)
distEuclidean.append(d)
distEuclidean.sort(reverse=True)
sumMin += distEuclidean.pop()
distEuclidean.clear()
igdVal = sumMin / len(truePF)
return igdVal
- Badges
- Science topic
- Similar topics
- Computer Science
- Program