params_to_non_central_logp

Import in python using from mcutils.scripts.MDE.fit_dispersion import params_to_non_central_logp

mcutils.scripts.MDE.fit_dispersion.params_to_non_central_logp(params, observed, bmat_list, include_S0=False, derivative=False, ncoils=1, fixed_noise_var=None)[source]

Computes the log(p) of the Non-central-chi fit to the data

Parameters
  • params – (4 + 4/5 * N, ) array for N b-shells (last parameter is the log of the noise variance, for others see params_to_signal)

  • observed – (N, ) list of array with the observed signal attenuations in each b-shell

  • bmat_list – (N, ) list of lists containing the b-matrices

  • include_S0 – include a parameter representing S0 rather than letting it be submerged in MD

  • derivative – also return the (3 + 4/5 * N, ) array with the error derivative

  • ncoils – number of coils (set to 1 for the Rician noise distribution)

  • fixed_noise_var – noise variance (if not set, last parameters is log(noise variance)

Returns

float with the log(p)