FitModel

FitModel#

class jetset.model_manager.FitModel(elec_distr=None, jet=None, name='no_name', out_dir=None, flag=None, template=None, loglog_poly=None, analytical=None, nu_size=200, cosmo=None, composite_expr=None, **keywords)[source]#

Bases: Model

Composite spectral model used for fitting observational datasets.

Notes

Combines one or more component models (for example, jets and templates), exposes a unified parameter interface, and evaluates either summed components or user-defined composite expressions.

Attributes Summary

composite_expr

Composite expr.

Methods Summary

add_component(m)

Add component.

del_component(m)

Del component.

eval([nu, fill_SED, get_model, loglog, ...])

Evaluate model output.

free(model, par_name)

Free.

free_all()

Set all model parameters to free state.

freeze(model, par_name)

Freeze.

freeze_all()

Freeze all model parameters.

get(model, par_name, *args)

Get.

get_par_by_name(model, par_name)

Return par by name.

link_par(par_name, from_model, to_model)

Link par.

load_model(file_name_or_obj[, from_string])

Load object state from disk.

plot_model([plot_obj, clean, sed_data, ...])

Plot the composite model, optional components, and residuals.

sed_tables_dict([restframe])

Sed tables dict.

set(model, par_name, *args, **kw)

Set.

set_fit_range([down_tol, up_tol])

Set fit range.

set_nu_grid([nu_min, nu_max, nu_size])

Set nu grid.

set_num_c_threads([num_c_threads, verbose])

Set C-backend thread count on all compatible model components.

set_par(model, par_name, val)

Set par.

show_model()

Print full composite-model summary and component details.

show_model_components()

Print a concise overview of component models.

Attributes Documentation

composite_expr#

Composite expr.

Returns:

Requested value.

Return type:

object

Methods Documentation

add_component(m)[source]#

Add component.

Parameters:

m (object) – Model/component object.

del_component(m)[source]#

Del component.

Parameters:

m (object) – Model/component object.

eval(nu=None, fill_SED=True, get_model=False, loglog=False, label=None, phys_output=False)[source]#

Evaluate model output.

Parameters:

  • nu (object, optional) – Frequency values in Hz.

  • fill_SED (bool, optional) – If True, store evaluated values into SED containers.

  • get_model (bool, optional) – If True, return model values.

  • loglog (bool, optional) – If True, operate in log10 space.

  • label (object, optional) – Label used in output or plots.

  • phys_output (bool, optional) – If True, return physical-output quantities when available.

Returns:

Computed value.

Return type:

object

free(model, par_name)[source]#

Free.

Parameters:

  • model (object) – Model instance.

  • par_name (object) – Parameter name.

free_all()[source]#

Set all model parameters to free state.

Notes

Delegates to CompositeModelParameterArray.free_all.

freeze(model, par_name)[source]#

Freeze.

Parameters:

  • model (object) – Model instance.

  • par_name (object) – Parameter name.

freeze_all()[source]#

Freeze all model parameters.

Notes

Delegates to CompositeModelParameterArray.freeze_all.

get(model, par_name, *args)[source]#

Get.

Parameters:

  • model (object) – Model instance.

  • par_name (object) – Parameter name.

  • *args (tuple) – Additional positional arguments.

get_par_by_name(model, par_name)[source]#

Return par by name.

Parameters:

  • model (object) – Model instance.

  • par_name (object) – Parameter name.

Returns:

Requested value.

Return type:

object

Link par.

Parameters:

  • par_name (object) – Parameter name.

  • from_model (object) – Source model/component for parameter linking.

  • to_model (object) – Destination model/component for parameter linking.

classmethod load_model(file_name_or_obj, from_string=False)[source]#

Load object state from disk.

Parameters:

  • file_name_or_obj (object) – Serialized model path or already-open object.

  • from_string (bool, optional) – If True, deserialize from an in-memory string payload.

Returns:

Loaded object.

Return type:

object

plot_model(plot_obj=None, clean=False, sed_data=None, frame='obs', skip_components=False, label=None, skip_sub_components=False, only_components=False, density=False)[source]#

Plot the composite model, optional components, and residuals.

Parameters:

  • plot_obj (PlotSED, optional) – Existing plot object to update. If None, a new one is created.

  • clean (bool, optional) – If True, clear previously drawn model lines before plotting.

  • sed_data (ObsData, optional) – Observational SED data. Used to initialize the plot helper and to compute residuals. If None, residual points are not added.

  • frame ({'obs', 'src'}, optional) – Frame used to display model curves.

  • skip_components (bool, optional) – If True, do not plot top-level component models.

  • label (str, optional) – Label for the total composite model curve. Defaults to self.name.

  • skip_sub_components (bool, optional) – If True, do not plot sub-components for each component model.

  • only_components (bool, optional) – If True, skip plotting the total composite curve (self.SED) and plot only component/sub-component curves (subject to skip flags).

  • density (bool, optional) – Passed to the plotting helper when creating a new plot object.

Returns:

Plot object with model curves (and residuals when sed_data is provided).

Return type:

PlotSED

sed_tables_dict(restframe='obs')[source]#

Sed tables dict.

Parameters:

restframe (str, optional) – Target rest frame for output.

Returns:

Computed value.

Return type:

object

set(model, par_name, *args, **kw)[source]#

Set.

Parameters:

  • model (object) – Model instance.

  • par_name (object) – Parameter name.

  • *args (tuple) – Additional positional arguments.

  • **kw (dict) – Additional keyword-value mapping.

set_fit_range(down_tol=0.1, up_tol=100)[source]#

Set fit range.

Parameters:

  • down_tol (float, optional) – Lower tolerance for parameter scans/intervals.

  • up_tol (int, optional) – Upper tolerance for parameter scans/intervals.

set_nu_grid(nu_min=None, nu_max=None, nu_size=None)[source]#

Set nu grid.

Parameters:

  • nu_min (object, optional) – Minimum frequency in Hz.

  • nu_max (object, optional) – Maximum frequency in Hz.

  • nu_size (object, optional) – Number of points for frequency grids.

set_num_c_threads(num_c_threads=None, verbose=False)[source]#

Set C-backend thread count on all compatible model components.

Parameters:

  • num_c_threads (int, optional) – Thread count forwarded unchanged to each component exposing set_num_c_threads.

  • verbose (bool, optional) – Verbosity flag forwarded to each component set_num_c_threads call.

Notes

Components without a set_num_c_threads method are skipped. Validation of num_c_threads is delegated to each component implementation. num_c_threads None, will trigger per-component automatic setup, such as multiprocessing.cpu_count().

set_par(model, par_name, val)[source]#

Set par.

Parameters:

  • model (object) – Model instance.

  • par_name (object) – Parameter name.

  • val (object) – Value to assign.

show_model()[source]#

Print full composite-model summary and component details.

Notes

Includes the per-component show_model output.

show_model_components()[source]#

Print a concise overview of component models.

Notes

This view does not print individual parameter tables.