tuskitoo.SpectralExtraction.fitting module

make_fit(ydata, num_source=2, initial_center=None, initial_separation=None, bound_sigma=None, fix_sep=None, fix_height=None, custom_expr=None, weights=None, param_limit=None, param_fix=None, param_value=None, verbose=False, distribution='gaussian')[source]

Fits a model to the provided data using either a Gaussian or Moffat profile.

Parameters:

ydataarray-like

The dependent data (observations or measurements) to be fitted by the model.

num_sourceint, optional, default=2

Number of sources (or components) to include in the fitting model.

initial_centerfloat, int, or None, optional, default=None

Initial estimate for the center (mean or peak position) of each source/component.

initial_separationlist or None, optional, default=None

Initial guess for the separation between the sources/components.

bound_sigmaarray-like or None, optional, default=None

Bounds or constraints on the sigma (standard deviation) for components beyond the first.

fix_seplist, bool, or None, optional, default=None

List of source indices for which the separation is fixed (i.e., not varied during fitting).

fix_heightbool or None, optional, default=None

Flag to fix the height (amplitude) of the sources/components.

custom_exprdict or None, optional, default=None

Dictionary mapping parameter names to custom expressions for the fitting function. Refer to https://lmfit.github.io/lmfit-py/index.html for details.

weightsarray-like or None, optional, default=None

Weights for each data point in ydata; useful for weighted fitting.

param_limitdict or None, optional, default=None

Dictionary specifying limits (min and max) for parameters. Example: {‘sigma_1’: (0.1, 5.0), ‘center_1’: (100, 200)}

param_fixlist or None, optional, default=None

List of parameter names to be fixed at a specified value during fitting.

param_valuedict or None, optional, default=None

Dictionary of initial parameter values. Example: {‘height_1’: 10.0, ‘sigma_1’: 2.0}

verbosebool, optional, default=False

If True, prints detailed information during the fitting process.

distributionstr, optional, default=”gaussian”

Type of distribution to use for fitting. Supported options are “gaussian” and “moffat”.

Returns:

resultlmfit.model.ModelResult

The result of the fitting process, including fitted parameters and fit statistics.

Raises:

ValueError:

If the provided distribution is not among the supported types.

Parameters:

ydata (array)

parallel_fit(image, error, num_source, pixel_limit=None, n_cpu=None, mask_list=[], **kwargs)[source]

Perform parallel fitting on a 2D image using multiple sources.

Parameters:

image2D array-like

The image data to be fitted.

error2D array-like

The error (sigma) associated with the image data.

num_sourceint

Number of sources (or components) to include in the fitting model.

pixel_limitlist or tuple of two ints, optional, default=None

Column indices [start, end] that limit the processing region. If None, all columns are processed.

n_cpuint or None, optional, default=None

Number of CPU cores to use for parallel processing. Defaults to the total available cores.

wavelenghtlist, optional, default=[]

(Not used yet) Intended to hold wavelength information for each pixel.

mask_listlist, optional, default=[]

List of column ranges to mask (exclude) from fitting. Each element should be a two-element iterable (start, stop).

**kwargsdict
Additional keyword arguments to be passed to the make_fit function. These can include:

  • initial_center

  • initial_separation

  • bound_sigma

  • fix_sep

  • fix_height

  • custom_expr

  • weights

  • param_limit

  • param_fix

  • param_value

  • verbose

  • distribution

Returns:

spectral_extraction_resultsdict
A dictionary containing:

  • “value”: Fitted parameter values.

  • “std”: Parameter uncertainties.

  • “name_params”: Names of the fitted parameters.

  • “extra_params”: Extra fit statistics (e.g., chisqr, redchi, AIC, BIC, R-squared).

  • “normalized_image”: The normalized image used for fitting.

  • “normalize_matrix”: Normalization factors for each pixel column.

  • “num_source”: Number of sources used.

  • “distribution”: Distribution type used.

  • “mask”: The mask applied to the image.

  • “parameter_number”: Number of parameters per source.