def calc_marg_stats(self, *params):
"""
Calculate the marginal density and Jacobian vs. nonlinear parameters
and memoize the results.
"""
if (params == self.pars):
return self.Q, self.suf, self.Jac
#... Evaluate the standardized models on the data, first calling the setup
#... function for the data if it is present.
args = list(params)
args.append(self.absc)
args = tuple(args)
if (self.setup != None):
aux = self.setup(*args)
args = list(args)
args.append(aux)
args = tuple(args)
for a in range(self.M):
g = self.basis[a]
self.modvals[a,:] = g(*args) / self.sigs
#... Calculate the metric and various derived quantities. Store them
#... for possible future use before returning.
self.metric, self.L, self.Jac, self.proj, self.ampl, self.suf =\
vba.metricAnalysis(self.modvals, self.std_smpls)
self.Q = self.dsqr - self.suf
self.pars = params
self.Jac = 1. / self.Jac
def calc_marg_stats(self, *params):
"""
Calculate the marginal density and Jacobian vs. nonlinear parameters
and memoize the results.
"""
if (params == self.pars):
return self.Q, self.suf, self.Jac
#... Evaluate the standardized models on the data, first calling the setup
#... function.
self.obj.set_nonlin(*params)
for a in range(self.M):
self.modvals[a,:] = self.obj.std_basis[a]()
#... Calculate the metric and various derived quantities. Store them
#... for possible future use before returning.
self.metric, self.L, self.Jac, self.proj, self.ampl, self.suf =\
vba.metricAnalysis(self.modvals, self.obj.std_smpls)
self.Q = self.dsqr - self.suf
self.pars = params
self.Jac = 1. / self.Jac
