Main routines¶
Main routines are the heart of spexxy, describing a specific process that should be run on the data.
There is one pre-defined main routine in spexxy:
ParamsFittakes one or more spectrum components and finds the best parameter combination to match a given spectrum.
MainRoutine¶
FilesRoutine¶
-
class
spexxy.main.FilesRoutine(*args, **kwargs)¶ FilesRoutine is the base class for all routines fitting files.
-
__call__(filename: str) → List[float]¶ Start the routine on the given file.
Parameters: filename – Name of file to process. Returns: List of final values of parameters, ordered in the same way as the return value of parameters()
-
__init__(*args, **kwargs)¶ Initialize a new MainRoutine object
-
columns() → List[str]¶ Get list of columns returned by __call__.
The returned list shoud include the list from parameters().
Returns: List of columns returned by __call__.
-
parameters() → List[str]¶ Get list of parameters fitted by this routine.
Returns: List of parameter names (including prefix) fitted by this routine.
-
ParamsFit¶
-
class
spexxy.main.ParamsFit(components: List[spexxy.component.component.Component] = None, tellurics: spexxy.component.component.Component = None, masks: List[spexxy.mask.mask.Mask] = None, weights: List[spexxy.weight.weight.Weight] = None, fixparams: List[str] = None, poly_degree: int = 40, maxfev: int = 500, ftol: float = 1.49012e-08, xtol: float = 1.49012e-08, factor: float = 100.0, epsfcn: float = 1e-07, min_valid_pixels: float = 0.5, plot_iterations: bool = False, *args, **kwargs)¶ ParamsFit is a fitting routine for spexxy that uses a Levenberg-Marquardt optimization to fit a set of model spectra (components) to a given spectrum.
-
__call__(filename: str) → List[float]¶ Start the fitting procedure on the given file.
Parameters: filename – Name of file to fit. Returns: List of final values of parameters, ordered in the same way as the return value of parameters()
-
__init__(components: List[spexxy.component.component.Component] = None, tellurics: spexxy.component.component.Component = None, masks: List[spexxy.mask.mask.Mask] = None, weights: List[spexxy.weight.weight.Weight] = None, fixparams: List[str] = None, poly_degree: int = 40, maxfev: int = 500, ftol: float = 1.49012e-08, xtol: float = 1.49012e-08, factor: float = 100.0, epsfcn: float = 1e-07, min_valid_pixels: float = 0.5, plot_iterations: bool = False, *args, **kwargs)¶ Initialize a new ParamsFit object
Parameters: - components – List of components (or descriptions to create them) to fit to the spectrum
- tellurics – Tellurics component to add to the fit.
- masks – List of mask objects to use for masking the spectrum.
- weights – List of weight objects to use for creating weights for the spectrum.
- fixparams – List of names of parameters to fix during the fit.
- poly_degree – Number of coefficients for the multiplicative polynomial.
- maxfev – The maximum number of calls to the function (see scipy documentation).
- ftol – Relative error desired in the sum of squares (see scipy documentation).
- xtol – Relative error desired in the approximate solution (see scipy documentation).
- factor – A parameter determining the initial step bound (factor * || diag * x||) (see scipy documentation).
- epsfcn – A variable used in determining a suitable step length for the forward- difference approximation of the Jacobian (see scipy documentation)
- min_valid_pixels – Fraction of minimum number of required pixels to continue with fit.
- plot_iterations – Plot all iterations into a PDF file.
-
columns() → List[str]¶ Get list of columns returned by __call__.
The returned list shoud include the list from parameters().
Returns: List of columns returned by __call__.
-
components¶ Returns all components used in this fit.
Returns: List of all components.
-
fit_parameters() → List[str]¶ Get list of parameters fitted by this routine. Exclude fixed parameters.
Returns: List of parameter names (including prefix) fitted by this routine.
-
parameters() → List[str]¶ Get list of parameters fitted by this routine.
Returns: List of parameter names (including prefix) fitted by this routine.
-
MultiMain¶
-
class
spexxy.main.MultiMain(routines: List[T] = None, iterations: int = None, max_iterations: int = 8, threshold: Dict[str, float] = None, poly_degree: int = 40, damped: bool = True, factors: list = (0.7, 0.3), *args, **kwargs)¶ MultiRun iterates over a given set of main routines and runs them sequentially.
-
__call__(filename: str) → List[float]¶ Process the given file.
Parameters: filename – Name of file to process. Returns: List of final values of parameters, ordered in the same way as the return value of parameters()
-
__init__(routines: List[T] = None, iterations: int = None, max_iterations: int = 8, threshold: Dict[str, float] = None, poly_degree: int = 40, damped: bool = True, factors: list = (0.7, 0.3), *args, **kwargs)¶ Initialize a new MultiMain object
Parameters: - routines – List of main routines to run.
- iterations – Number of iterations for the whole cycle.
- max_iterations – If set to a value >=2, the fit runs until it converges or until the number of iterations reaches max_iterations.
- threshold – Dictionary that contains the absolute values for each fit parameter below which the fit is considered as converged.
- poly_degree – Degree of Legendre polynomial used for the continuum fit.
- damped – If True the fit is repeated with a damping factor in case the fit does not converge.
- factors – List of damping factors used if the fit does not converge.
-
columns() → List[str]¶ Get list of columns returned by __call__.
The returned list should include the list from parameters().
Returns: List of columns returned by __call__.
-
convergence(results)¶ Returns true if the fit satisfies the convergence criteria for all fit parameters.
-
parameters() → List[str]¶ Get list of parameters fitted by this routine.
Returns: List of parameter names (including prefix) fitted by this routine.
-