In this tutorial we will use the scikit-geodesic package to compute a geodesic in an isotropic Riemannian manifold with coefficient exp(-<n,x>). The script is available in the example directory of the code.
# An Example Script Illustrating how to find the geodesic for an isotropic Riemannian manifold with metric coefficient
# exp(-<n,x>) where n is a constant vector.
from math import exp
import numpy as np
from geodesic.geometry import Curve
from geodesic.curve_shorten import compute_geodesic
from multiprocessing import cpu_count
# Set dimension of the problem
dimension = 4
# Set parameters for computation
number_of_global_nodes = 16
number_of_local_nodes = 8
maximum_average_node_movement = 0.001
number_of_cpu = cpu_count()
# Create start and end point NumPy arrays
start_point = np.zeros(dimension)
start_point[0] = -1
end_point = np.zeros(dimension)
end_point[0] = 1
# Create constant vector n
alpha = 0.65
n = alpha*np.ones(dimension)
n[0] = 0
# Define function to describe metric coefficient
def metric_coefficient(x):
return exp(-np.inner(n,x))
print 'Starting Example Calculation...'
# Create curve object for calculation
curve = Curve(start_point, end_point, number_of_global_nodes)
# Apply curve shortening procedure to minimise length
compute_geodesic(curve, number_of_local_nodes, maximum_average_node_movement, metric_coefficient, number_of_cpu)
# Print shortened curve points
print curve.get_points()