LinearFitResult

Definition

LinearFitResult is an immutable dataclass with slots=True, used as the return object of lin_fit.

Purpose

Collect the fit parameters, associated uncertainties, residuals, diagnostics, and optional figure in a single typed container.

Fields

• slope: slope of the fitted line.

• intercept: intercept of the fitted line.

• slope_std: standard uncertainty on the slope.

• intercept_std: standard uncertainty on the intercept.

• covariance: covariance between the two fit parameters.

• correlation: correlation coefficient between slope and intercept.

• residuals: vector of residuals y - (m x + c).

• residual_std: compact estimate of residual dispersion.

• chi2: chi-squared of the fit.

• reduced_chi2: reduced chi-squared.

• dof: degrees of freedom, equal to n - 2.

• iterations: number of weight updates performed.

• converged: indicates whether the iteration with sigma_x satisfied the stopping criterion.

• figure: matplotlib object or None when show_plot=False.

When to read it

Use this object when you want to:

• retrieve the fit parameters without having to parse a string or a legend

• check fit quality through residuals and reduced_chi2

• decide whether to save or reuse the generated figure

Example

result = lin_fit(x, y, sigma_y, show_plot=False)

print(result.slope)
print(result.intercept_std)
print(result.reduced_chi2)

Notes

The class does not perform calculations on its own: it is a representation of the output produced by lin_fit.