Plotting, recipes, models, and solvers

Plotting

qsospec.plot_line_result(result, output_path=None, ax=None, show_components=True, title=None)

Plot one fitted local line-complex result.

Returns the Matplotlib axes when output_path is not provided, otherwise saves the figure and returns the written path.

Parameters:

• result (FitResult)

• output_path (str | None)

• show_components (bool)

• title (str | None)

qsospec.plot_local_result(result, output_path=None, show_components=True)

Plot all windows from a local fit in stacked panels.

Parameters:

• result (LocalFitResult)

• output_path (str | None)

• show_components (bool)

qsospec.save_local_window_plots(result, output_dir, show_components=True)

Write one PNG per local-fit window and return {window: path}.

Parameters:

• result (LocalFitResult)

• output_dir (str)

• show_components (bool)

Return type:

Dict[str, str]

Registries

qsospec.recipes.list_complexes()

Return type:

Tuple[ComplexRecipe, …]

qsospec.recipes.get(value)

Parameters:

value (str)

Return type:

ComplexRecipe

qsospec.recipes.resolve(value)

Parameters:

value (str)

Return type:

str

qsospec.recipes.describe(value)

Parameters:

value (str)

Return type:

Dict[str, Any]

qsospec.recipes.generic_narrow_lines(line_ids, **changes)

Parameters:

• line_ids (Iterable[str])

• changes (Any)

Return type:

ComplexRecipe

qsospec.lines.list()

Return all registered line definitions in wavelength order.

Return type:

Tuple[LineDefinition, …]

qsospec.lines.get(value)

Return one immutable line definition.

Parameters:

value (str)

Return type:

LineDefinition

qsospec.lines.resolve(value)

Resolve a canonical line ID from a normalized or historical alias.

Parameters:

value (str)

Return type:

str

Models and solvers

qsospec.models.gaussian(wave, amp, center, sigma)

Evaluate amp * exp(-0.5 * ((wave - center) / sigma)**2).

Parameters:

• wave (ndarray)

• amp (float)

• center (float)

• sigma (float)

Return type:

ndarray

qsospec.models.gaussian_partials(wave, amp, center, sigma)

Return dense model derivatives with columns amp, center, sigma.

Parameters:

• wave (ndarray)

• amp (float)

• center (float)

• sigma (float)

Return type:

ndarray

qsospec.models.lorentzian(wave, amp, center, gamma)

Evaluate a peak-normalized Lorentzian profile.

Parameters:

• wave (ndarray)

• amp (float)

• center (float)

• gamma (float)

Return type:

ndarray

qsospec.models.lorentzian_partials(wave, amp, center, gamma)

Return dense derivatives with columns amp, center, gamma.

Parameters:

• wave (ndarray)

• amp (float)

• center (float)

• gamma (float)

Return type:

ndarray

qsospec.models.continuum(wave, mode, params, wave_ref, scale)

Evaluate the MVP local continuum model.

Parameters:

• wave (ndarray)

• mode (str | None)

• params (ndarray)

• wave_ref (float)

• scale (float)

Return type:

ndarray

qsospec.models.continuum_partials(wave, mode, wave_ref, scale)

Return dense model derivatives for the local continuum parameters.

Parameters:

• wave (ndarray)

• mode (str | None)

• wave_ref (float)

• scale (float)

Return type:

ndarray

qsospec.solvers.solve_variable_projection(flux, err, nonlinear_initial, nonlinear_bounds, linear_bounds, evaluator, *, jacobian_method='semi_analytic', max_nfev=None)

Solve a bounded separable nonlinear least-squares problem.

Parameters:

• flux (ndarray)

• err (ndarray)

• nonlinear_initial (ndarray)

• nonlinear_bounds (Tuple[ndarray, ndarray])

• linear_bounds (Tuple[ndarray, ndarray])

• evaluator (Callable[[ndarray, bool], Tuple[ndarray, Sequence[ndarray] | None]])

• jacobian_method (str)

• max_nfev (int | None)

Return type:

VariableProjectionResult

qsospec.solvers.run_least_squares(packed, wave, flux, err, jacobian='analytic_dense', max_nfev=None)

Run SciPy least-squares for a packed local line-complex model.

Parameters:

• packed (PackedParameters)

• wave (ndarray)

• flux (ndarray)

• err (ndarray)

• jacobian (str)

• max_nfev (int | None)