fit_static_thy

TRXASprefitpack.driver.fit_static_thy(thy_peak: Sequence[ndarray], fwhm_G_init: float, fwhm_L_init: float, policy: str, peak_shift: Optional[ndarray] = None, peak_scale: Optional[ndarray] = None, edge: Optional[str] = None, edge_pos_init: Optional[ndarray] = None, edge_fwhm_init: Optional[ndarray] = None, base_order: Optional[int] = None, method_glb: Optional[str] = None, method_lsq: Optional[str] = 'trf', kwargs_glb: Optional[dict] = None, kwargs_lsq: Optional[dict] = None, bound_fwhm_G: Optional[Tuple[float, float]] = None, bound_fwhm_L: Optional[Tuple[float, float]] = None, bound_peak_shift: Optional[Sequence[Tuple[float, float]]] = None, bound_peak_scale: Optional[Sequence[Tuple[float, float]]] = None, bound_edge_pos: Optional[Sequence[Tuple[float, float]]] = None, bound_edge_fwhm: Optional[Sequence[Tuple[float, float]]] = None, e: Optional[ndarray] = None, intensity: Optional[ndarray] = None, eps: Optional[ndarray] = None) → StaticResult[source]

driver routine for fitting static spectrum with sum of voigt broadend thoretical spectrum, edge and polynomial base line. To solve least square optimization problem efficiently, it implements the seperation scheme. Moreover this driver uses two step algorithm to search best parameter, its covariance and estimated parameter error.

Step 1. (method_glb) Use global optimization to find rough global minimum of our objective function. In this stage, it use analytic gradient for scalar residual function.

Step 2. (method_lsq) Use least squares optimization algorithm to refine global minimum of objective function and approximate covariance matrix. Because of linear and non-linear seperation scheme, the analytic jacobian for vector residual function is hard to optain. Thus, in this stage, it uses numerical jacobian.

Parameters

  • thy_peak (sequence of np.ndarray) – peak position and intensity for theoretically calculated spectrum

  • fwhm_G_init (float) – initial gaussian part of fwhm parameter

  • fwhm_L_init (float) – initial lorenzian part of fwhm parameter

  • policy ({'shift', 'scale', 'both'}) – policy to match discrepancy between thoretical spectrum and experimental one

  • peak_shift (np.ndarray) – peak shift parameter for each species

  • peak_scale (np.ndarray) – peak scale parameter for each species

  • edge ({'g', 'l'}) – type of edge function. If edge is not set, edge feature is not included.

  • edge_pos_init (np.ndarray) – initial edge position

  • edge_fwhm_init (np.ndarray) – initial fwhm parameter of edge

  • method_glb ({None, 'basinhopping', 'ampgo'}) – Method for global optimization

  • method_lsq ({'trf', 'dogbox', 'lm'}) – method of local optimization for least_squares minimization (refinement of global optimization solution)

  • kwargs_glb – keyward arguments for global optimization solver

  • kwargs_lsq – keyward arguments for least square optimization solver

  • bound_fwhm_G (tuple) – boundary for fwhm_G parameter. If bound_fwhm_G is None, the upper and lower bound are given as (fwhm_G/2, 2*fwhm_G).

  • bound_fwhm_L (tuple) – boundary for fwhm_L parameter. If bound_fwhm_L is None, the upper and lower bound are given as (fwhm_L/2, 2*fwhm_L).

  • bound_peak_shift (sequence of tuple) – boundary for peak shift parameter. If bound_peak_shift is None, the upper and lower bound are given by set_bound_e0.

  • bound_peak_scale (sequence of tuple) – boundary for peak scale parameter. If bound_peak_scale is None, the upper and lower bound are given as (0.9*peak_scale, 1.1*peak_scale).

  • bound_edge_pos (sequence of tuple) – boundary for edge position, if bound_edge_pos is None and edge is set, the upper and lower bound are given by set_bound_t0.

  • bound_edge_fwhm (sequence of tuple) – boundary for fwhm parameter of edge feature. If bound_edge_fwhm is None, the upper and lower bound are given as (edge_fwhm/2, 2*edge_fwhm).

  • e (np.narray) – energy range for data

  • intensity (np.ndarray) – intensity of static spectrum data

  • eps (np.ndarray) – estimated errors of static spectrum data

  • Returns – StaticResult class object

  • Note –

    • if initial fwhm_G is zero then such voigt component is treated as lorenzian component

    • if initial fwhm_L is zero then such voigt component is treated as gaussian component

    • Every theoretical spectrum is normalize.