def __init__(self, target_eigenvalue, init_state_vect, dgl_coefficients, domain):
if not all([isinstance(state, (int, float, long)) for state in init_state_vect]) \
and len(init_state_vect) == 4 and isinstance(init_state_vect, (list, tuple)):
raise TypeError
if not len(dgl_coefficients) == 3 and isinstance(dgl_coefficients, (list, tuple)) \
and all([callable(coef) or isinstance(coef, (int, float, long)) for coef in dgl_coefficients]):
raise TypeError
if not isinstance(domain, (np.ndarray, list)) \
or not all([isinstance(num, (int, long, float)) for num in domain]):
raise TypeError
if isinstance(target_eigenvalue, complex):
self._eig_val_real = target_eigenvalue.real
self._eig_val_imag = target_eigenvalue.imag
elif isinstance(target_eigenvalue, (int, long, float)):
self._eig_val_real = target_eigenvalue
self._eig_val_imag = 0.
else:
raise TypeError
self._init_state_vect = init_state_vect
self._a2, self._a1, self._a0 = [ut._convert_to_function(coef) for coef in dgl_coefficients]
self._domain = domain
state_vect = self._transform_eigenfunction()
self._transf_eig_func_real, self._transf_d_eig_func_real = state_vect[0:2]
self._transf_eig_func_imag, self._transf_d_eig_func_imag = state_vect[2:4]
Function.__init__(self, self._phi, nonzero=(domain[0], domain[-1]), derivative_handles=[self._d_phi])
def __init__(self, omega, param, spatial_domain, norm_fac=1.):
self._omega = omega
self._param = param
self.norm_fac = norm_fac
a2, a1, a0, _, _ = self._param
self._eta = -a1 / 2. / a2
Function.__init__(self, self._phi, nonzero=spatial_domain, derivative_handles=[self._d_phi, self._dd_phi])
def __init__(self, function, M, b, l, scale_func=None, nested_lambda=False):
if not callable(function):
raise TypeError
if not isinstance(M, np.ndarray) or len(M.shape) != 2 or np.diff(M.shape) != 0 or M.shape[0]%1 != 0:
raise TypeError
if not all([isinstance(num, (int, long, float)) for num in [b, l]]):
raise TypeError
self.function = function
self.M = M
self.b = b
self.l = l
if scale_func == None:
self.scale_func = lambda z: 1
else:
self.scale_func = scale_func
self.n = int(M.shape[0]/2)
self.l0 = l/self.n
self.z_disc = np.array([(i+1)*self.l0 for i in range(self.n)])
if not nested_lambda:
# iteration mode
Function.__init__(self,
self._call_transformed_func,
nonzero=(0, l),
derivative_handles=[])
else:
# nested lambda mode
self.x_func_vec = list()
for i in range(self.n):
self.x_func_vec.append(AddMulFunction(
partial(lambda z, k: self.scale_func(k*self.l0 + z)*self.function(k*self.l0 + z), k=i)))
for i in range(self.n):
self.x_func_vec.append(AddMulFunction(
partial(lambda z, k: self.scale_func(self.l - k*self.l0 - z)*self.function(self.l - k*self.l0 - z), k=i)))
self.y_func_vec = np.dot(self.x_func_vec, np.transpose(M))
Function.__init__(self,
self._call_transformed_func_vec,
nonzero=(0, l),
derivative_handles=[])
def __init__(self, om, param, spatial_domain, phi_0=1):
self._om = om
self._param = param
self.phi_0 = phi_0
Function.__init__(self, self._phi, nonzero=spatial_domain, derivative_handles=[self._d_phi, self._dd_phi])
