Crate polyfit_residuals

Efficiently compute the residual errors for all possible polynomial models up to some degree for given data.

Example

For examples please have a look at the exported functions like residuals_from_front.

Modules

• poly
  Basic polynomials in a Newton basis.
• weighted
  Provides versions of the main functions for the case of a weighted least squares fit. So we calculate the optimal target values of min_{p polynomial of deg d} ∑ᵢ wᵢ(p(xᵢ) - yᵢ)² for all d and valid discrete intervals of i.

Structs

• PolyFit
  A fit polynomial together with its residual error

Enums

• FitError
  The different errors that can occur during the polynomial fitting process.

Functions

• all_residuals
  Compute the residual squared errors (RSS) for all polynomials of degree at most max_deg for the data segments xs[j..=i], ys[j..=i] for all i, j.
• all_residuals_par
  A parallel version of all_residuals_par. Please have a look at the sequential version for details.
• residuals_from_front
  Compute the residual squared errors (RSS) for all polynomials of degree at most max_deg for the data segments xs[0..=i], ys[0..=i] for all i.
• solve_upper_triangular_system
  Solves the linear system matrix_product(lhs, x) = rhs for x.
• try_fit_poly
  Try fitting a polynomial to some data.
• try_fit_poly_with_residual
  Try fitting a polynomial to some data and also compute the residual error.