Note

Go to the end to download the full example code

Generate low discrepancy sequencesΒΆ

In this examples we are going to expose the available low discrepancy sequences in order to approximate some integrals.

The following low-discrepancy sequences are available:

  • Sobol

  • Faure

  • Halton

  • reverse Halton

  • Haselgrove

To illustrate these sequences we generate their first 1024 points and compare with the sequence obtained from the pseudo random generator (Merse Twister) as the latter has a higher discrepancy.

import openturns as ot
import openturns.viewer as viewer
from matplotlib import pylab as plt

ot.Log.Show(ot.Log.NONE)

  1. Sobol sequence

dimension = 2
size = 1024
sequence = ot.SobolSequence(dimension)
sample = sequence.generate(size)
graph = ot.Graph("Sobol", "", "", True, "")
cloud = ot.Cloud(sample)
graph.add(cloud)
view = viewer.View(graph)
Sobol

  1. Halton sequence

dimension = 2
sequence = ot.HaltonSequence(dimension)
sample = sequence.generate(size)
graph = ot.Graph("Halton", "", "", True, "")
cloud = ot.Cloud(sample)
graph.add(cloud)
view = viewer.View(graph)
Halton

  1. Halton sequence in high dimension: bad filling in upper dimensions

dimension = 20
sequence = ot.HaltonSequence(dimension)
sample = sequence.generate(size).getMarginal([dimension - 2, dimension - 1])
graph = ot.Graph(
    "Halton (" + str(dimension - 2) + "," + str(dimension - 1) + ")",
    "dim " + str(dimension - 2),
    "dim " + str(dimension - 1),
    True,
    "",
)
cloud = ot.Cloud(sample)
graph.add(cloud)
view = viewer.View(graph)
Halton (18,19)

  1. Scrambled Halton sequence in high dimension

dimension = 20
sequence = ot.HaltonSequence(dimension)
sequence.setScrambling("RANDOM")
sample = sequence.generate(size).getMarginal([dimension - 2, dimension - 1])
graph = ot.Graph(
    "Halton (" + str(dimension - 2) + "," + str(dimension - 1) + ")",
    "dim " + str(dimension - 2),
    "dim " + str(dimension - 1),
    True,
    "",
)
cloud = ot.Cloud(sample)
graph.add(cloud)
view = viewer.View(graph)
Halton (18,19)

  1. Reverse Halton sequence

dimension = 2
sequence = ot.ReverseHaltonSequence(dimension)
sample = sequence.generate(size)
print(
    "discrepancy=",
    ot.LowDiscrepancySequenceImplementation.ComputeStarDiscrepancy(sample),
)
graph = ot.Graph("Reverse Halton", "", "", True, "")
cloud = ot.Cloud(sample)
graph.add(cloud)
view = viewer.View(graph)
Reverse Halton
discrepancy= 0.0035074981424325635

  1. Haselgrove sequence

dimension = 2
sequence = ot.HaselgroveSequence(dimension)
sample = sequence.generate(size)
graph = ot.Graph("Haselgrove", "", "", True, "")
cloud = ot.Cloud(sample)
graph.add(cloud)
view = viewer.View(graph)
Haselgrove

Compare with uniform random sequence

distribution = ot.ComposedDistribution([ot.Uniform(0.0, 1.0)] * 2)
sample = distribution.getSample(size)
print(
    "discrepancy=",
    ot.LowDiscrepancySequenceImplementation.ComputeStarDiscrepancy(sample),
)
graph = ot.Graph("Mersenne Twister", "", "", True, "")
cloud = ot.Cloud(sample)
graph.add(cloud)
view = viewer.View(graph)
plt.show()
Mersenne Twister
discrepancy= 0.03921823278089642