sidpy.proc.fitter.SidFitter

class sidpy.proc.fitter.SidFitter(sidpy_dataset, fit_fn, xvec=None, ind_dims=None, guess_fn=None, num_fit_parms=None, km_guess=False, n_clus=None, return_cov=False, return_std=False, return_fit=False, fit_parameter_labels=None, num_workers=2, threads=2)[source]

Bases: object

Parameters:

  • sidpy_dataset ((sidpy.Dataset) Sidpy dataset object to be fit)

  • fit_fn ((function) Function used for fitting.)

  • arguments. (Should take xvec as the first argument and parameters as the rest of the)

  • xvec ((numpy ndarray or list of numpy ndarrays) (Optional))

  • xvec

  • array (Independent variable for fitting. Should be an)

  • provided (If the guess_fn is NOT)

  • xvecs (the dimension arrays are assumed to be)

  • ind_dims ((tuple) (Optional) Tuple with integer entries of the dimensions) – over which to parallelize. These should be the independent variable for the fitting. If NOT provided, it is assumed that all the non-spectral dimensions are independent dimensions.

  • guess_fn ((function) (optional) This optional function should be utilized to generate priors for the full fit)

  • (xvec (It takes)

  • parameters. (yvec) as inputs and should return the fit)

  • provided

  • num_fit_parms. (then the user MUST input the)

  • num_fit_parms ((int) Number of fitting parameters. This is needed IF the guess function is not provided to set)

  • function. (the priors for the parameters for the curve_fit)

  • km_guess ((bool) (default False) When set to True: Divides the spectra into clusters using)

  • sklearn.optimize.kMeans

  • centers (applies the fitting function on the cluster)

:param : :param uses the results as priors to each spectrum of the cluster.: :param n_clus: :type n_clus: (int) (default None) Used only when km_guess is set to True. Determines the number of clusters to be :param formed for sklearn.optimize.kmeans. If not provided then n_clus = self.num_computations/100: :param return_std: :type return_std: (bool) (default False) Returns the dataset with estimated standard deviation of the parameter :param values. Square roots of the diagonal of the covariance matrix.: :param return_cov: :type return_cov: (bool) (default False) Returns the estimated covariance of fitting parameters. Confer :param scipy.optimize.curve_fit for further details: :param return_fit: when set to true :type return_fit: (bool) (default False) Returns the fitted sidpy dataset using the optimal parameters :param fit_parameter_labels: :type fit_parameter_labels: (list) (default None) List of parameter labels :param num_workers: :type num_workers: (int) (default =2) Number of workers to use when setting up Dask client :param threads: :type threads: (int) (default =2) Number of threads to use when setting up Dask client :param Returns: :param ——-: :param sidpy.dataset: :type sidpy.dataset: if return_cov and return_fit are both set to False :param List: :type List: containing sidpy.dataset objects, if either of return_cov or return fit is set to True :param If multiple datasets are expected: :param the order of the returned datasets is: :param [sidpy.dataset with mean parameter values: :param : :param sidpy.dataset with estimated covariances of the fitting parameters: :param : :param sidpy.dataset that is fit with the parameters obtained after fitting]:

Methods

default_curve_fit

do_fit

Perform the fit. **kwargs: extra parameters passed to scipy.optimize.curve_fit, e.g. bounds, type of lsq algorithm, etc.

do_guess

If a guess_fn is provided: Applies the guess_fn to get priors for the fitting parameters.

get_fitted_dataset

This method returns the fitted dataset using the parameters generated by the fit function

get_km_priors

visualize_fit_results

Calls the interactive visualizer for comparing raw and fit datasets.
do_fit(**kwargs)[source]

Perform the fit. **kwargs: extra parameters passed to scipy.optimize.curve_fit, e.g. bounds, type of lsq algorithm, etc.

do_guess()[source]

If a guess_fn is provided: Applies the guess_fn to get priors for the fitting parameters. self.prior is set as the output of guess function at each of the ind_dims

Returns: None ——-

get_fitted_dataset()[source]

This method returns the fitted dataset using the parameters generated by the fit function

visualize_fit_results(figure=None, horizontal=True)[source]

Calls the interactive visualizer for comparing raw and fit datasets.

Inputs:

  • figure: (Optional, default None) - handle to existing figure

  • horiziontal: (Optional, default True) - whether spectrum should be plotted horizontally