def test_initialisation():
I, J = 2, 3
R = numpy.ones((I, J))
M = numpy.ones((I, J))
K = 4
# Init ones
init_UV = 'ones'
nmf = NMF(R, M, K)
nmf.initialise(init_UV)
assert numpy.array_equal(numpy.ones((2, 4)), nmf.U)
assert numpy.array_equal(numpy.ones((3, 4)), nmf.V)
# Init random
init_UV = 'random'
nmf = NMF(R, M, K)
nmf.initialise(init_UV)
for (i, k) in itertools.product(range(0, I), range(0, K)):
assert nmf.U[i, k] > 0 and nmf.U[i, k] < 1
for (j, k) in itertools.product(range(0, J), range(0, K)):
assert nmf.V[j, k] > 0 and nmf.V[j, k] < 1
# We now run the VB algorithm on each of the M's for each fraction.
all_performances = {metric: [] for metric in metrics}
average_performances = {metric: []
for metric in metrics} # averaged over repeats
for (fraction, Ms, Ms_test) in zip(fractions_unknown, all_Ms, all_Ms_test):
print "Trying fraction %s." % fraction
# Run the algorithm <repeats> times and store all the performances
for metric in metrics:
all_performances[metric].append([])
for (repeat, M, M_test) in zip(range(0, repeats), Ms, Ms_test):
print "Repeat %s of fraction %s." % (repeat + 1, fraction)
nmf = NMF(R, M, K)
nmf.initialise(init_UV, expo_prior)
nmf.run(iterations)
# Measure the performances
performances = nmf.predict(M_test)
for metric in metrics:
# Add this metric's performance to the list of <repeat> performances for this fraction
all_performances[metric][-1].append(performances[metric])
# Compute the average across attempts
for metric in metrics:
average_performances[metric].append(
sum(all_performances[metric][-1]) / repeats)
print "repeats=%s \nfractions_unknown = %s \nall_performances = %s \naverage_performances = %s" % \