Source code for sigima.proc.signal.fitting
# -*- coding: utf-8 -*-
#
# Licensed under the terms of the BSD 3-Clause
# (see sigima/LICENSE for details)
"""
Curve fitting operations
========================
This module provides curve fitting operations for signal objects:
- Linear and polynomial fits
- Gaussian, Lorentzian, and Voigt fits
- Exponential and CDF fits
.. note::
Most operations use functions from :mod:`sigima.tools.signal.fitting` for
actual computations.
"""
from __future__ import annotations
from typing import Callable
import guidata.dataset as gds
import numpy as np
from sigima.config import _
from sigima.objects import SignalObj
from sigima.proc.base import dst_2_to_1
from sigima.proc.decorator import computation_function
from sigima.tools.signal import fitting
from .base import dst_1_to_1
def __generic_fit(
src: SignalObj,
fitfunc: Callable[[np.ndarray, np.ndarray], tuple[np.ndarray, dict[str, float]]],
) -> SignalObj:
"""Generic fitting function.
Args:
src: source signal
fitfunc: fitting function
Returns:
Fitting result signal object
"""
dst = dst_1_to_1(src, fitfunc.__name__)
# Fit only on ROI if available
x_roi = src.x[~src.get_masked_view().mask]
y_roi = src.get_masked_view().compressed()
_fitted_y_roi, fit_params = fitfunc(x_roi, y_roi)
# Evaluate fit on full x range
fitted_y = fitting.evaluate_fit(src.x, **fit_params)
dst.set_xydata(src.x, fitted_y)
# Store fit parameters in metadata
dst.metadata["fit_params"] = fit_params
return dst
[docs]
@computation_function()
def linear_fit(src: SignalObj) -> SignalObj:
"""Compute linear fit with :py:func:`numpy.polyfit`
Args:
src: source signal
Returns:
Result signal object
"""
return __generic_fit(src, fitting.linear_fit)
[docs]
class PolynomialFitParam(gds.DataSet, title=_("Polynomial fit")):
"""Polynomial fitting parameters"""
degree = gds.IntItem(_("Degree"), 3, min=1, max=10, slider=True)
[docs]
@computation_function()
def polynomial_fit(src: SignalObj, p: PolynomialFitParam) -> SignalObj:
"""Compute polynomial fit with :py:func:`numpy.polyfit`
Args:
src: source signal
p: polynomial fitting parameters
Returns:
Result signal object
"""
# Note: no need to check degree here as gds.IntItem already enforces min=1
return __generic_fit(src, lambda x, y: fitting.polynomial_fit(x, y, p.degree))
[docs]
@computation_function()
def gaussian_fit(src: SignalObj) -> SignalObj:
"""Compute Gaussian fit with :py:func:`scipy.optimize.curve_fit`
Args:
src: source signal
Returns:
Result signal object
"""
return __generic_fit(src, fitting.gaussian_fit)
[docs]
@computation_function()
def lorentzian_fit(src: SignalObj) -> SignalObj:
"""Compute Lorentzian fit with :py:func:`scipy.optimize.curve_fit`
Args:
src: source signal
Returns:
Result signal object
"""
return __generic_fit(src, fitting.lorentzian_fit)
[docs]
@computation_function()
def voigt_fit(src: SignalObj) -> SignalObj:
"""Compute Voigt fit with :py:func:`scipy.optimize.curve_fit`
Args:
src: source signal
Returns:
Result signal object
"""
return __generic_fit(src, fitting.voigt_fit)
[docs]
@computation_function()
def exponential_fit(src: SignalObj) -> SignalObj:
"""Compute exponential fit with :py:func:`scipy.optimize.curve_fit`
Args:
src: source signal
Returns:
Result signal object
"""
return __generic_fit(src, fitting.exponential_fit)
[docs]
@computation_function()
def cdf_fit(src: SignalObj) -> SignalObj:
"""Compute CDF fit with :py:func:`scipy.optimize.curve_fit`
Args:
src: source signal
Returns:
Result signal object
"""
return __generic_fit(src, fitting.cdf_fit)
[docs]
@computation_function()
def planckian_fit(src: SignalObj) -> SignalObj:
"""Compute Planckian fit with :py:func:`scipy.optimize.curve_fit`
Args:
src: source signal
Returns:
Result signal object
"""
return __generic_fit(src, fitting.planckian_fit)
[docs]
@computation_function()
def twohalfgaussian_fit(src: SignalObj) -> SignalObj:
"""Compute two-half-Gaussian fit with :py:func:`scipy.optimize.curve_fit`
Args:
src: source signal
Returns:
Result signal object
"""
return __generic_fit(src, fitting.twohalfgaussian_fit)
[docs]
@computation_function()
def sigmoid_fit(src: SignalObj) -> SignalObj:
"""Compute sigmoid fit with :py:func:`scipy.optimize.curve_fit`
Args:
src: source signal
Returns:
Result signal object
"""
return __generic_fit(src, fitting.sigmoid_fit)
[docs]
@computation_function()
def piecewiseexponential_fit(src: SignalObj) -> SignalObj:
"""Compute piecewise exponential fit (raise-decay) with
:py:func:`scipy.optimize.curve_fit`
Args:
src: source signal
Returns:
Result signal object
"""
return __generic_fit(src, fitting.piecewiseexponential_fit)
[docs]
@computation_function()
def sinusoidal_fit(src: SignalObj) -> SignalObj:
"""Compute sinusoidal fit with :py:func:`scipy.optimize.curve_fit`
Args:
src: source signal
Returns:
Result signal object
"""
return __generic_fit(src, fitting.sinusoidal_fit)
[docs]
@computation_function()
def evaluate_fit(src1: SignalObj, src2: SignalObj) -> SignalObj:
"""Evaluate fit function from src1 on the x-axis of src2.
This function extracts fit parameters from `src1` (which must contain fit metadata
from a previous fitting operation) and evaluates the fit function on the x-axis
of `src2`.
Args:
src1: Signal object containing fit parameters in metadata (from a fit operation)
src2: Signal object whose x-axis will be used for evaluation
Returns:
New signal with the fit evaluated on src2's x-axis
"""
fit_params = extract_fit_params(src1)
dst = dst_2_to_1(src1, src2, "evaluate_fit")
# Evaluate fit on src2's x-axis
x = src2.x
y = fitting.evaluate_fit(x, **fit_params)
dst.set_xydata(x, y)
dst.title = f"Fitted {fit_params['fit_type']}"
# Copy fit parameters to destination metadata
dst.metadata["fit_params"] = fit_params
return dst
