def test_to_explicit_vectors():
print "\n-- 'to_explicit_bool_vector', 'to_explicit_list' --"
fileNameX = join(data_directory, 'Torus_X.csv')
X, _, _ = load_X(fileNameX, n=8, zeroindexing=False)
print "Torus X:\n", X
Xb = to_explicit_bool_vector(X)
print 'Xb:\n', Xb
Xl = to_explicit_list(X)
print 'Xl:\n', Xl
Y = np.array([[0, 0, 0], [0, 0, 1], [0, 1, 1], [0, 0, -1], [0, 0, 0.0001],
[0, 0, 0.001]])
print "\nY:\n", Y
Yb = to_explicit_bool_vector(Y)
print 'Yb:\n', Yb
Yl = to_explicit_list(Y)
print 'Yl:\n', Yl
def test_load_and_save_X():
print "\n-- 'load_X', 'save_X'"
print "Load matrix and specify n=8"
# filename1 = 'data/Torus_X.csv'
X, _, _ = load_X(join(data_directory, 'Torus_X.csv'),
n=8,
zeroindexing=False)
print "X:\n", X
print "Shape:", X.shape
print "\nSave and load matrix with n=8 k=3"
filename2 = 'Torus_X2.csv'
save_X(join(data_directory, filename2), X)
X, _, _ = load_X(join(data_directory, filename2), n=8, k=3)
print "X:\n", X
print "\nSave and load matrix with n=8 k=3, delimiter= ' ', format gz"
filename3 = 'Torus_X3.gz'
save_X(join(data_directory, filename3), X, delimiter=' ')
X, _, _ = load_X(join(data_directory, filename3), n=8, k=3, delimiter=None)
print "X:\n", X
print "\nInput contains float"
X, _, _ = load_X(join(data_directory, 'Torus_X4.csv'), n=8, k=3)
print "X:\n", X
print "\nInput contains only 1 (explicit beliefs)"
X, _, _ = load_X(join(data_directory, 'Torus_X5.csv'), n=8, k=3)
print "X:\n", X
print "\nInput contains only two columns. Correct"
X, _, _ = load_X(join(data_directory, 'Torus_X_twoColumns.csv'),
n=8,
k=3,
delimiter=None)
print "X:\n", X
print "\nInput contains only two columns. With assert error (node with several classes)"
try:
X, _, _ = load_X(join(data_directory,
'Torus_X_twoColumns_assertError.csv'),
n=8,
k=3,
delimiter=None)
except AssertionError, e:
print "! Assertion error:\n", e
def test_to_explicit_vectors():
print "\n-- 'to_explicit_bool_vector', 'to_explicit_list' --"
fileNameX = join(data_directory, 'Torus_X.csv')
X, _, _ = load_X(fileNameX, n=8, zeroindexing=False)
print "Torus X:\n", X
Xb = to_explicit_bool_vector(X)
print 'Xb:\n', Xb
Xl = to_explicit_list(X)
print 'Xl:\n', Xl
Y = np.array( [[0,0,0],
[0,0,1],
[0,1,1],
[0,0,-1],
[0,0,0.0001],
[0,0,0.001]])
print "\nY:\n", Y
Yb = to_explicit_bool_vector(Y)
print 'Yb:\n', Yb
Yl = to_explicit_list(Y)
print 'Yl:\n', Yl
def test_load_and_save_X():
print "\n-- 'load_X', 'save_X'"
print "Load matrix and specify n=8"
# filename1 = 'data/Torus_X.csv'
X, _, _ = load_X(join(data_directory, 'Torus_X.csv'), n=8, zeroindexing=False)
print "X:\n", X
print "Shape:", X.shape
print "\nSave and load matrix with n=8 k=3"
filename2 = 'Torus_X2.csv'
save_X(join(data_directory, filename2), X)
X, _, _ = load_X(join(data_directory, filename2), n=8, k=3)
print "X:\n", X
print "\nSave and load matrix with n=8 k=3, delimiter= ' ', format gz"
filename3 = 'Torus_X3.gz'
save_X(join(data_directory, filename3), X, delimiter=' ')
X, _, _ = load_X(join(data_directory, filename3), n=8, k=3, delimiter=None)
print "X:\n", X
print "\nInput contains float"
X, _, _ = load_X(join(data_directory, 'Torus_X4.csv'), n=8, k=3)
print "X:\n", X
print "\nInput contains only 1 (explicit beliefs)"
X, _, _ = load_X(join(data_directory, 'Torus_X5.csv'), n=8, k=3)
print "X:\n", X
print "\nInput contains only two columns. Correct"
X, _, _ = load_X(join(data_directory, 'Torus_X_twoColumns.csv'), n=8, k=3, delimiter=None)
print "X:\n", X
print "\nInput contains only two columns. With assert error (node with several classes)"
try:
X, _, _ = load_X(join(data_directory, 'Torus_X_twoColumns_assertError.csv'), n=8, k=3, delimiter=None)
except AssertionError, e:
print "! Assertion error:\n", e
def test_transform_beliefs():
print "\n-- 'check_normalized_beliefs', 'to_centering_beliefs' --"
X = np.array([[1.0001, 0, 0]])
print "X: ", X
assert check_normalized_beliefs(X)
print "X centered: ", to_centering_beliefs(X)
Y = np.array([0.9999, 0, 0])
print "Y: ", Y
assert check_normalized_beliefs(Y)
print "Y centered: ", to_centering_beliefs(Y)
Z = np.array([[1.001, 0, 0]])
print "Z: ", Z
assert not check_normalized_beliefs(Z)
W = np.array([0.999, 0, 0])
print "W: ", W
assert not check_normalized_beliefs(W)
print "\n-- 'check_centered_beliefs', 'from_centering_beliefs'"
Xc = np.array([[1.0001, -1, 0]])
print "Xc: ", Xc
assert check_centered_beliefs(Xc)
print "Xc uncentered: ", from_centering_beliefs(Xc)
Yc = np.array([0.9999, -1, 0])
print "Yc: ", Yc
assert check_centered_beliefs(Yc)
print "Yc uncentered: ", from_centering_beliefs(Yc)
Zc = np.array([[1.001, -1, 0]])
print "Zc: ", Zc
assert not check_centered_beliefs(Zc)
Wc = np.array([0.999, -1, 0])
print "Wc: ", Wc
assert not check_centered_beliefs(Wc)
print "\n--'to_centering_beliefs', 'from_centering_beliefs' for matrices --"
X = np.array( [[1,0,0],
[0.8,0.2,0],
[1./3,1./3,1./3],
[0,0,1],
[0,0,1],
[0.5,0,0.5]])
print "X original:\n", X
print "np.sum(X,1):\n", np.sum(X,1)
print "X.sum(axis=1, keepdims=True):\n", X.sum(axis=1, keepdims=True)
print "X.shape:", X.shape
print "len(X.shape): ", len(X.shape)
Xc = to_centering_beliefs(X)
print "X centered:\n", Xc
Y = from_centering_beliefs(Xc)
print "X again un-centered:\n", Y
fileNameX = join(data_directory, 'Torus_X.csv')
X, _, _ = load_X(fileNameX, n=8, zeroindexing=False)
X = X.dot(0.1)
print "\nCentered X for Torus example as input\n", X
Xc = from_centering_beliefs(X)
print "X un-centered:\n", Xc
X = np.array( [[1,0,0]])
print "\nX original:\n", X
Xc = to_centering_beliefs(X)
print "X centered:\n", Xc
Y = from_centering_beliefs(Xc)
print "X back non-centered:\n", Y
X = np.array( [1,0,0])
print "\nX original:\n", X
print "np.sum(X,0):", np.sum(X,0)
print "X.sum(axis=0, keepdims=True):", X.sum(axis=0, keepdims=True)
print "X.shape: ", X.shape
print "len(X.shape): ", len(X.shape)
save_W(join(data_directory, '{}_{}_W.csv'.format(filename, n)),
W,
saveWeights=False)
save_X(join(data_directory, '{}_{}_X.csv'.format(filename, n)), X0)
save_tuple(n, 'graph', time_graph)
else:
W, _ = load_W(join(data_directory,
'{}_{}_W.csv'.format(filename, n)),
skiprows=1,
zeroindexing=True,
n=None,
doubleUndirected=False)
X0, _, _ = load_X(join(data_directory,
'{}_{}_X.csv'.format(filename, n)),
n=None,
k=None,
skiprows=1,
zeroindexing=True)
# -- Repeat loop
for i in range(repeat):
print("\n repeat: {}".format(i))
X2, ind = replace_fraction_of_rows(X0,
1 - f,
avoidNeighbors=avoidNeighbors,
W=W)
# -- Estimate Esp_max
start = time.time()
eps_max = eps_convergence_directed_linbp(P=H0,
def test_transform_beliefs():
print "\n-- 'check_normalized_beliefs', 'to_centering_beliefs' --"
X = np.array([[1.0001, 0, 0]])
print "X: ", X
assert check_normalized_beliefs(X)
print "X centered: ", to_centering_beliefs(X)
Y = np.array([0.9999, 0, 0])
print "Y: ", Y
assert check_normalized_beliefs(Y)
print "Y centered: ", to_centering_beliefs(Y)
Z = np.array([[1.001, 0, 0]])
print "Z: ", Z
assert not check_normalized_beliefs(Z)
W = np.array([0.999, 0, 0])
print "W: ", W
assert not check_normalized_beliefs(W)
print "\n-- 'check_centered_beliefs', 'from_centering_beliefs'"
Xc = np.array([[1.0001, -1, 0]])
print "Xc: ", Xc
assert check_centered_beliefs(Xc)
print "Xc uncentered: ", from_centering_beliefs(Xc)
Yc = np.array([0.9999, -1, 0])
print "Yc: ", Yc
assert check_centered_beliefs(Yc)
print "Yc uncentered: ", from_centering_beliefs(Yc)
Zc = np.array([[1.001, -1, 0]])
print "Zc: ", Zc
assert not check_centered_beliefs(Zc)
Wc = np.array([0.999, -1, 0])
print "Wc: ", Wc
assert not check_centered_beliefs(Wc)
print "\n--'to_centering_beliefs', 'from_centering_beliefs' for matrices --"
X = np.array([[1, 0, 0], [0.8, 0.2, 0], [1. / 3, 1. / 3, 1. / 3],
[0, 0, 1], [0, 0, 1], [0.5, 0, 0.5]])
print "X original:\n", X
print "np.sum(X,1):\n", np.sum(X, 1)
print "X.sum(axis=1, keepdims=True):\n", X.sum(axis=1, keepdims=True)
print "X.shape:", X.shape
print "len(X.shape): ", len(X.shape)
Xc = to_centering_beliefs(X)
print "X centered:\n", Xc
Y = from_centering_beliefs(Xc)
print "X again un-centered:\n", Y
fileNameX = join(data_directory, 'Torus_X.csv')
X, _, _ = load_X(fileNameX, n=8, zeroindexing=False)
X = X.dot(0.1)
print "\nCentered X for Torus example as input\n", X
Xc = from_centering_beliefs(X)
print "X un-centered:\n", Xc
X = np.array([[1, 0, 0]])
print "\nX original:\n", X
Xc = to_centering_beliefs(X)
print "X centered:\n", Xc
Y = from_centering_beliefs(Xc)
print "X back non-centered:\n", Y
X = np.array([1, 0, 0])
print "\nX original:\n", X
print "np.sum(X,0):", np.sum(X, 0)
print "X.sum(axis=0, keepdims=True):", X.sum(axis=0, keepdims=True)
print "X.shape: ", X.shape
print "len(X.shape): ", len(X.shape)
