MATH3230
Numerical analysis
Error and stability, root finding, direct and iterative linear solvers, least squares, interpolation and uniform approximation, numerical differentiation, quadrature and eigenvalue iterations.
Your progress
0/ 29 sections marked complete
Next section: Floating point, conditioning, and stability
Completion marks are saved in this browser.
This course in the web
In the MATH3230 notes
- Theorem cards
- 55
- Definitions
- 45
- Counterexamples
- 31
- Concept lenses
- 17
- Histories of ideas
- 0
- Proof X-rays
- 29
No concept lens from this course is recorded in the web yet.
Counts cover the 29 public sections of this course and what is written in them, not recorded connections. 0 of these sections are indexed into the web in detail so far.
In the webTheorem cards 55 · Definitions 45 · Counterexamples 31 · Concept lenses 17 · Proof X-rays 29
No concept lens from this course is recorded in the web yet.
Counts cover the 29 public sections of this course and what is written in them, not recorded connections. 0 of these sections are indexed into the web in detail so far.
Chapter 1 · 1 section
Error and stability
Representation, sensitivity and numerical stability.
Theorems 2 · Definitions 4Counterexamples 1Lenses 1
Chapter outline →See this chapter on the web →Not yet indexed in the web
- 1.1
Floating point, conditioning, and stability
Distinguish problem conditioning from algorithmic error.
Chapter 2 · 5 sections
Root finding
Brackets and local iteration.
Theorems 7 · Definitions 2Counterexamples 5Lenses 1Proof X-rays 5
Chapter outline →See this chapter on the web →Not yet indexed in the web
- 2.1
Bracketing and Newton's method
Prove bisection and Newton convergence under stated hypotheses.
- 2.2
Safeguarded secant steps and bisection
Combine secant proposals with bisection and prove a bracket-width guarantee.
- 2.3
Designing fixed-point iterations and bounding their error
Verify an iteration on an interval and turn successive steps into an error certificate.
- 2.4
Newton's method at a multiple root
Derive the loss of quadratic convergence and a multiplicity correction.
- 2.5
Newton's method for nonlinear systems
Extend Newton's iteration to systems F(x)=0 by solving a linear system with the Jacobian, prove local quadratic convergence, and examine singular Jacobians and starting points.
Chapter 3 · 3 sections
Linear systems
Elimination, residuals and sensitivity.
Theorems 5 · Definitions 3Counterexamples 3Lenses 1Proof X-rays 3
Chapter outline →See this chapter on the web →Not yet indexed in the web
- 3.1
Elimination, residuals, and conditioning
Solve with pivoting and interpret residual and condition number.
- 3.2
Cholesky factorization for positive definite systems
Derive a symmetric factorization, solve a system, and identify the limits of its guarantees.
- 3.3
Tridiagonal elimination and positive definite systems
Derive a linear-cost direct solver, prove when its pivots exist, and distinguish structure from numerical accuracy.
Chapter 4 · 1 section
Approximation from sampled data
Polynomial interpolation, central differences, and composite trapezoids with explicit error models.
Theorems 2 · Definitions 3Counterexamples 1Lenses 1Proof X-rays 1
Chapter outline →See this chapter on the web →Not yet indexed in the web
- 4.1
Interpolants, differences, and quadrature
Prove interpolation remainders, quantify sample sensitivity, and derive differentiation and quadrature bounds.
Chapter 5 · 3 sections
Interpolation beyond existence
Incremental forms, evaluation, conditioning, and extremal node products.
Theorems 6 · Definitions 6Counterexamples 5Lenses 2Proof X-rays 4
Chapter outline →See this chapter on the web →Not yet indexed in the web
- 5.1
Newton form, barycentric evaluation, and node choice
Prove Newton and barycentric identities, the Lebesgue inequality, and Chebyshev node-product minimax behavior.
- 5.2
Hermite data and natural cubic splines
Interpolate derivative data with precise remainders, then construct the unique natural cubic spline and prove its minimum-bending property.
- 5.3
Piecewise-linear interpolation on a mesh
Build a local interpolant on uneven nodes and distinguish approximation, derivative, and data errors.
Chapter 6 · 5 sections
Quadrature and error estimation
Symmetry, orthogonality, polynomial precision, and conditional adaptive estimates.
Theorems 9 · Definitions 1Counterexamples 5Lenses 5Proof X-rays 5
Chapter outline →See this chapter on the web →Not yet indexed in the web
- 6.1
Gaussian quadrature and adaptive error models
Derive midpoint and Simpson bounds, prove Gaussian precision, and explain Richardson and adaptive stopping limitations.
- 6.3
Weighted Gaussian rules: nodes, weights and limits
Construct weighted Gaussian quadrature, prove optimal polynomial precision and positive weights, and assess its limits.
- 6.4
Gaussian quadrature: a derivative error certificate
Prove the weighted Gaussian derivative remainder, compute its sharp constant, and derive a composite two-node error certificate.
- 6.5
Refining trapezoids and building a Romberg table
Reuse samples, cancel leading errors, and distinguish extrapolation from certification.
- 6.6
Adaptive Simpson: an estimate and a certificate
Build an adaptive algorithm, derive its estimator, and distinguish model-based stopping from a derivative-bound guarantee.
Chapter 7 · 4 sections
Iterative linear solvers
Spectral iteration, Krylov energy minimization, and positive definite preconditioning.
Theorems 8 · Definitions 8Counterexamples 4Lenses 1Proof X-rays 4
Chapter outline →See this chapter on the web →Not yet indexed in the web
- 7.1
Stationary iteration, Krylov spaces, and preconditioning
Prove stationary spectral convergence and conjugate-gradient minimization, bounds, and preconditioning identities.
- 7.2
SOR: relaxation and energy descent
Derive SOR and prove convergence for symmetric positive-definite systems using energy descent.
- 7.3
Gauss–Seidel certificates beyond symmetry
Prove a row-by-row convergence certificate for Gauss–Seidel on nonsymmetric systems.
- 7.4
Stopping an iteration with rounded sweeps
Certify iteration stopping with contraction and explicit sweep-error bounds.
Chapter 8 · 2 sections
Error budgets and reliable arithmetic
Fourth-order differentiation, uncertainty tradeoffs, and summation stability.
Theorems 4 · Definitions 4Counterexamples 2Lenses 2Proof X-rays 2
Chapter outline →See this chapter on the web →Not yet indexed in the web
- 8.1
Extrapolation and reliable arithmetic
Prove a sharp fourth-order derivative bound and componentwise summation stability, then inspect compensation and its limits.
- 8.2
Quadratic roots, cancellation, and conditioning
Separate cancellation in quadratic formulas from discriminant and root sensitivity.
Chapter 13 · 1 section
Numerical differentiation
Construct differentiation stencils and reason about their sensitivity to data.
Theorems 2 · Definitions 2Counterexamples 1Proof X-rays 1
Chapter outline →See this chapter on the web →Not yet indexed in the web
- 13.1
One-sided differentiation on unequal nodes
Construct derivative weights from samples on one side of a point at unequal spacings, and bound truncation and data-perturbation errors separately.
Chapter 14 · 1 section
Reliable factorizations and least squares
Pivot growth, factor and solve backward errors, and orthogonal least-squares reduction.
Theorems 3 · Definitions 3Counterexamples 1Lenses 1Proof X-rays 1
Chapter outline →See this chapter on the web →Not yet indexed in the web
- 14.1
Pivot growth, Householder QR, and least squares
Connect factorization errors to the solved problem, and derive least squares through orthogonal transformations.
Chapter 15 · 1 section
Least squares without full column rank
Separate fitted vectors, nullspace freedom, minimum norm, and rank decisions through SVD.
Theorems 2 · Definitions 3Counterexamples 1Lenses 1Proof X-rays 1
Chapter outline →See this chapter on the web →Not yet indexed in the web
- 15.1
SVD, minimum-norm least squares, and numerical rank
Resolve nonuniqueness by orthogonal geometry, prove the pseudoinverse formula, and distinguish exact rank from a thresholded model.
Chapter 16 · 1 section
Best uniform approximation
Existence, uniqueness, alternation, and explicit Chebyshev minimax approximants.
Theorems 2 · Definitions 3Counterexamples 1Lenses 1Proof X-rays 1
Chapter outline →See this chapter on the web →Not yet indexed in the web
- 16.1
Best polynomials and equioscillation
Prove existence, uniqueness, and the alternation characterization of best uniform polynomial approximation.
Chapter 17 · 1 section
Eigenvalue iterations
Compute dominant and shifted eigenpairs by repeated multiplication and solves, with convergence rates and failure cases.
Theorems 3 · Definitions 3Counterexamples 1Proof X-rays 1
Chapter outline →See this chapter on the web →Not yet indexed in the web
- 17.1
The power method and inverse iteration
Compute a dominant or a selected eigenpair by repeated multiplication or shifted solves, prove the convergence rates, and identify the hypotheses that make them fail.