Skip to article frontmatterSkip to article content
Site not loading correctly?

This may be due to an incorrect BASE_URL configuration. See the MyST Documentation for reference.

B-splines

Owner import path

jaxstro.numerics.splines

Purpose

B-spline basis construction, evaluation, calculus, fitting, and tensor-product design mechanics for fixed knot vectors.

Public records and callables

BSpline1D, open_uniform_knots, adaptive_open_uniform_knots, bspline_basis, bspline_design_matrix, bspline_eval, bspline_eval_deboor, bspline_derivative, bspline_antiderivative, bspline_integral, bspline_roughness_penalty, fit_bspline_lstsq, and tensor_product_design_matrix.

Shape and dtype expectations

Knots and coefficients are one-dimensional floating arrays along the spline axis. Degree and axis are static; tensor-product designs require aligned sample rows.

JAX transforms and AD classification

Fixed-knot evaluation composes with JIT and smooth-pathwise AD away from knots. Adaptive knot placement and rank or branch changes are preprocessing boundaries.

Failure behavior

Invalid degree, knot order, coefficient shape, or unsupported axis raises for concrete inputs. Fitting exposes the underlying linear-solve behavior.

See B-splines and Validation.

Canonical import example

from jaxstro.numerics.splines import BSpline1D