beta = np.random.rand(N)
theta = 10
# Simple testing of the performance of the Python and Scipy implementations
import time
t0 = time.time()
rbf_network(X, beta, theta)
print("Python: ", time.time() - t0)
t0 = time.time()
rbf_scipy(X, beta)
print("Scipy: ", time.time() - t0)
# Testing the performance of Cython
t0 = time.time()
rbf_network_cython(X, beta, theta)
print("Cython: ", time.time() - t0)
# Cython implementation of a Radial Basis Function (RBF) approximation scheme
#
# TODO: Write the Cython implementation in a separate fastloop.pyx file, compile and import it here
#
# from fastloop import rbf_network_cython
# Make up some data
D = 5
N = 1000
X = np.array([np.random.rand(N) for d in range(D)]).T
beta = np.random.rand(N)
theta = 10
# Simple testing of the performance of the Python and Scipy implementations
import time
# python
t0 = time.time()
rbf_network(X, beta, theta)
print("Python: ", time.time() - t0)
# scipy
t0 = time.time()
rbf_scipy(X, beta)
print("Scipy: ", time.time() - t0)
# Testing the performance of Cython
t0 = time.time()
fastloop.rbf_network_cython(X, beta, theta)
print("Cython: ", time.time() - t0)
def dummy(X, beta, theta):
rbf_network_cython(X, beta, theta)