funclp.Fit module

class funclp.Fit(function, estimator, **kwargs)[source]

Bases: ABC, CudaReference

Class defining fitting algorithms.

Parameters:: kwargs (dict) – Attributes to change.

function

function instance to fit

Type:: Function

estimator

estimator instance to use to fit

Type:: Estimator

fit

Core function used defining the fitting algorithm.

Type:: method

max_iterations

Maximum number of loops in the algorithm used.

Type:: int

property bool2fit

property convergence

static cpu_convergence(ftol, old_chi2, new_chi2, gtol, gradient, xtol, parameters, indices, parameters_steps, converged)[source]

property estimator

abstractmethod fit_init()[source]: Fit initialization specific to optimizer

abstractmethod fit_optimize()[source]: Fit main logic optimizer

ftol = 1e-08

property function

static gpu_convergence(ftol, old_chi2, new_chi2, gtol, gradient, xtol, parameters, indices, parameters_steps, converged)[source]

gtol = 0

property index2fit

property lower_bounds

max_iterations = 200

max_retries = 10

property name

property nparameters2fit

property parameters2fit

property upper_bounds

xtol = 1e-08

Optimizers