def Q_I_metric(point, g_function):
"""The function to calculate the inverse matrix Q.
Args:
point(numpy.array) :
Array of vector in dimension R^3N space. This vector is repasented as the points in curve.
g_function (func):
this function is given by user in k_g_function directionary.
Return :
Q_I_matrix (numpy.array.matrix):
Q_I_matrix is given by the certain point and function g(x). Q=g*g^T+\epsilon norm(0,0.001)
"""
g_matrix = g_function(point)
dimension = len(point)
Q_matrix = np.zeros([dimension, dimension])
for i in range(dimension):
for j in range(dimension):
for k in range(3):
Q_matrix[i][j] += g_matrix[k][i] * g_matrix[k][j]
Q_matrix[i][j] += np.random.normal(0, 0.001)
rank = np.linalg.matrix_rank(Q_matrix)
Q_I_matrix = linalg.inv(Q_matrix)
# c=np.linalg.det(Q_matrix)
return Q_I_matrix
def Q_I_metric(point,g_function):
"""The function to calculate the inverse matrix Q.
Args:
point(numpy.array) :
Array of vector in dimension R^3N space. This vector is repasented as the points in curve.
g_function (func):
this function is given by user in k_g_function directionary.
Return :
Q_I_matrix (numpy.array.matrix):
Q_I_matrix is given by the certain point and function g(x). Q=g*g^T+\epsilon norm(0,0.001)
"""
g_matrix=g_function(point)
dimension=len(point)
Q_matrix = np.zeros([dimension, dimension])
for i in range(dimension):
for j in range(dimension):
for k in range(3):
Q_matrix[i][j] += g_matrix[k][i] * g_matrix[k][j]
Q_matrix[i][j] += np.random.normal(0, 0.001)
rank = np.linalg.matrix_rank(Q_matrix)
Q_I_matrix = linalg.inv(Q_matrix)
# c=np.linalg.det(Q_matrix)
return Q_I_matrix
def gradient_function(point, k_function, g_function):
dimension = len(point)
gradient = np.zeros(dimension)
k_vector = k_function(point)
g_matrix = g_function(point)
# print len(k_function(point))
#m,n=g_matrix.shape
for i in range(dimension):
gradient[i] = k_vector[i]
for j in range(3):
gradient[i] += math.sqrt(
2.0 * 0.0001) * g_matrix[j][i] * np.random.normal(0, 0.001)
return gradient
def gradient_function(point,k_function,g_function):
dimension = len(point)
gradient = np.zeros(dimension)
k_vector = k_function(point)
g_matrix = g_function(point)
# print len(k_function(point))
#m,n=g_matrix.shape
for i in range(dimension):
gradient[i] = k_vector[i]
for j in range(3):
gradient[i] += math.sqrt(2.0*0.0001)*g_matrix[j][i]*np.random.normal(0, 0.001)
return gradient
