def test_predict():
R = numpy.array([[1, 2], [3, 4]], dtype=float)
M = numpy.array([[1, 1], [0, 1]])
(I, J, K, L) = (2, 2, 3, 1)
F = numpy.array([[1, 2, 3], [4, 5, 6]])
S = numpy.array([[7], [8], [9]])
G = numpy.array([[10], [11]])
#R_predicted = numpy.array([[500,550],[1220,1342]],dtype=float)
nmtf = NMTF(R, M, K, L)
nmtf.F = F
nmtf.S = S
nmtf.G = G
performances = nmtf.predict(M)
MSE = (499 * 499 + 548 * 548 + 1338 * 1338) / 3.0
R2 = 1 - (499**2 + 548**2 + 1338**2) / (
(4.0 / 3.0)**2 + (1.0 / 3.0)**2 + (5.0 / 3.0)**2) #mean=7.0/3.0
#mean_real=7.0/3.0,mean_pred=2392.0/3.0 -> diff_real=[-4.0/3.0,-1.0/3.0,5.0/3.0],diff_pred=[-892.0/3.0,-742.0/3.0,1634.0/3.0]
Rp = ((-4.0 / 3.0 * -892.0 / 3.0) + (-1.0 / 3.0 * -742.0 / 3.0) +
(5.0 / 3.0 * 1634.0 / 3.0)) / (math.sqrt((-4.0 / 3.0)**2 +
(-1.0 / 3.0)**2 +
(5.0 / 3.0)**2) *
math.sqrt((-892.0 / 3.0)**2 +
(-742.0 / 3.0)**2 +
(1634.0 / 3.0)**2))
assert performances['MSE'] == MSE
assert abs(performances['R^2'] - R2) < 0.000000001
assert abs(performances['Rp'] - Rp) < 0.000000001
def test_predict():
R = numpy.array([[1,2],[3,4]],dtype=float)
M = numpy.array([[1,1],[0,1]])
(I,J,K,L) = (2,2,3,1)
F = numpy.array([[1,2,3],[4,5,6]])
S = numpy.array([[7],[8],[9]])
G = numpy.array([[10],[11]])
#R_predicted = numpy.array([[500,550],[1220,1342]],dtype=float)
nmtf = NMTF(R,M,K,L)
nmtf.F = F
nmtf.S = S
nmtf.G = G
performances = nmtf.predict(M)
MSE = ( 499*499 + 548*548 + 1338*1338 ) / 3.0
R2 = 1 - (499**2 + 548**2 + 1338**2)/((4.0/3.0)**2 + (1.0/3.0)**2 + (5.0/3.0)**2) #mean=7.0/3.0
#mean_real=7.0/3.0,mean_pred=2392.0/3.0 -> diff_real=[-4.0/3.0,-1.0/3.0,5.0/3.0],diff_pred=[-892.0/3.0,-742.0/3.0,1634.0/3.0]
Rp = ((-4.0/3.0*-892.0/3.0)+(-1.0/3.0*-742.0/3.0)+(5.0/3.0*1634.0/3.0)) / (math.sqrt((-4.0/3.0)**2+(-1.0/3.0)**2+(5.0/3.0)**2) * math.sqrt((-892.0/3.0)**2+(-742.0/3.0)**2+(1634.0/3.0)**2))
assert performances['MSE'] == MSE
assert abs(performances['R^2'] - R2) < 0.000000001
assert abs(performances['Rp'] - Rp) < 0.000000001
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)
# Run the VB algorithm
nmtf = NMTF(R,M,K,L)
nmtf.initialise(init_S,init_FG)
nmtf.run(iterations)
# Measure the performances
performances = nmtf.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" % \
(repeats,fractions_unknown,all_performances,average_performances)
'''
repeats=10