fit module

File : fit.py Author : Rolf Verberg <rolfverberg AT gmail dot com> Description: General curve fitting module

class Component(model, prefix='')

Bases: object

A model fit component.

class Components

Bases: dict

The dictionary of model fit components.

add(model, prefix=''): Add a model to the model fit component dictionary.

property components: Return the model fit component dictionary.

class Fit(nxdata, config)

Bases: object

Wrapper class for scipy/lmfit.

add_model(model, prefix): Add a model component to the fit model.

add_parameter(parameter): Add a fit parameter to the fit model.

property best_errors: Return errors in the best fit parameters.

property best_fit: Return the best fit.

property best_parameters: Return the best fit parameters.

property best_values: Return values of the best fit parameters.

property chisqr: Return the chisqr value of the best fit.

property components: Return the fit model components info.

property covar: Return the covarience matrix of the best fit parameters.

eval(x, result=None): Evaluate the best fit.

fit(config=None, **kwargs): Fit the model to the input data.

static guess_init_peak(x, y, *args, center_guess=None, use_max_for_center=True): Return a guess for the initial height, center and fwhm for a single peak.

property init_parameters: Return the initial parameters for the fit model.

property init_values: Return the initial values for the fit parameters.

property normalization_offset: Return the normalization_offset for the fit model.

property num_func_eval: Return the number of function evaluations for the best fit.

property parameters: Return the fit parameter info.

plot(y=None, y_title=None, title=None, result=None, skip_init=False, plot_comp=True, plot_comp_legends=False, plot_residual=False, plot_masked_data=True, **kwargs): Plot the best fit.

print_fit_report(result=None, show_correl=False): Print a fit report.

property redchi: Return the redchi value of the best fit.

property residual: Return the residual in the best fit.

property success: Return the success value for the fit.

property var_names: Return the variable names for the covarience matrix property.

property x: Return the input x-array.

property y: Return the input y-array.

class FitMap(nxdata, config)

Bases: Fit

Wrapper to the Fit class to fit data on a N-dimensional map.

property best_errors: Return errors in the best fit parameters.

property best_fit: Return the best fits.

best_parameters(dims=None): Return the best fit parameters.

property best_values: Return values of the best fit parameters.

property chisqr: Return the chisqr value of each best fit.

property components: Return the fit model components info.

property covar: Return the covarience matrices of the best fit parameters.

fit(config=None, **kwargs): Fit the model to the input data.

freemem(): Free memory allocated for parallel processing.

property max_nfev: Return if the maximum number of function evaluations is reached for each fit.

property num_func_eval: Return the number of function evaluations for each best fit.

property out_of_bounds: Return the out_of_bounds value of each best fit.

plot(dims=None, y_title=None, plot_residual=False, plot_comp_legends=False, plot_masked_data=True, **kwargs): Plot the best fits.

property redchi: Return the redchi value of each best fit.

property residual: Return the residual in each best fit.

property success: Return the success value for each fit.

property var_names: Return the variable names for the covarience matrix property.

property y: Return the input y-array.

property ymap: Return the input y-array map.

class FitProcessor

Bases: Processor

A processor to perform a fit on a data set or data map.

static create_multipeak_model(model_config): Create a multipeak model.

process(data, config=None)

Fit the data and return a CHAP.utils.fit.Fit or CHAP.utils.fit.FitMap object depending on the dimensionality of the input data. The input data should contain a NeXus NXdata object, with properly defined signal and axis.

Parameters:

data (list[PipelineData]) – Input data containing the nexusformat.nexus.NXdata object to fit.
config (dict, optional) – Fit configuration.

Raises:

ValueError – Invalid input or configuration parameter.

Returns:

The fitted data object.

Return type:

Union[CHAP.utils.fit.Fit, CHAP.utils.fit.FitMap]

class ModelResult(model, parameters, x=None, y=None, method=None, ast=None, res_par_exprs=None, res_par_indices=None, res_par_names=None, result=None)

Bases: object

The result of a model fit, mimicking the functionality of a similarly named class in the lmfit library.

eval_components(x=None, parameters=None)

Evaluate each component of a composite model function.

Parameters:

x (Union[list, np.ndarray], optional) – Independent variable, defaults to None, in which case the class variable x is used.
parameters (Parameters, optional) – Composite model parameters, defaults to None, in which case the class variable params is used.

Returns:

A dictionary with component name and evealuated function values key, value pairs.

Return type:

dict

fit_report(show_correl=False)

Generates a report of the fitting results with their best parameter values and uncertainties.

Parameters:: show_correl (bool, optional) – Whether to show list of correlations, defaults to False.

class Parameters

Bases: dict

A dictionary of FitParameter objects, mimicking the functionality of a similarly named class in the lmfit library.

add(parameter, prefix='')

Add a fit parameter.

Parameters:

parameter (Union[str, FitParameter]) – The fit parameter to add to the dictionary.
prefix (str, optional) – The prefix for the model to which this parameter belongs, defaults to ‘’.