Fitter¶

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

Documentation

class mcutils.scripts.MDE.fit_dispersion.Fitter(data_list, bmat_list, include_S0=False, ncoils=0, noise_var=None)[source]¶

__init__(data_list, bmat_list, include_S0=False, ncoils=0, noise_var=None)[source]¶

Provides a fit for a single voxel

Parameters

data_list – (K,) length list with (M_k, ) array for K shells of N voxels with M_k acquisitions
bmat_list – (K,) length list with (M_k, 3, 3) array of b-matrices
include_S0 – if True include a parameter representing the S0 for every shell
ncoils – If set to non-zero: assumes a non-central model for the noise with 2 ncoils degrees of freedom rather than Gaussian
noise_var – Set to fix the noise variance rather than allow it to be a free parameters (only used if ncoils is non-zero)

Returns

best-fit parameters (same length as init_params)

Methods

`iterate`(init_params)	Iterates between fitting the individual shells and the angles
`only_angle`(init_params)	Fits only the angular parameters
`per_shell`(init_params)	Fits the parameters for every shell separately
`single`(init_params[, ignore_angle])	Provides a single fit to the full dataset at once

Attributes