|
CMP++: Uncertainty Quantification & Bayesian Calibration
|
Legendre orthogonal polynomial basis. More...
#include <poly.h>
Static Public Member Functions | |
| static double | evaluate (size_t deg, double xi) |
| static double | normSquared (size_t deg) |
| static Eigen::MatrixXd | getJacobiMatrix (int numPoints) |
| static double | getPDFWeight () |
| static double | mapToPhysical (double xi, double lowerBound, double upperBound) |
| static double | mapToCanonical (double x, double lowerBound, double upperBound) |
Legendre orthogonal polynomial basis.
Mathematical Formulation Legendre polynomials \(P_n(\xi)\) are orthogonal with respect to the uniform density function \(w(\xi) = 1/2\) on \([-1, 1]\). The recurrence relation is:
\[ (n+1) P_{n+1}(\xi) = (2n+1) \xi P_n(\xi) - n P_{n-1}(\xi) \]
with \(\int_{-1}^1 P_m(\xi) P_n(\xi) d\xi = \frac{2}{2n+1} \delta_{mn}\).
Implementation Algorithm
evaluate() computes polynomial values using Bonnet's recurrence relation.getJacobiMatrix() constructs the symmetric tridiagonal Jacobi matrix \(\mathbf{J}\) with \(\alpha_i = 0\) and off-diagonal \(\beta_i = \frac{i}{\sqrt{(2i-1)(2i+1)}}\).
|
inlinestatic |
|
inlinestatic |
|
inlinestatic |
|
inlinestatic |
|
inlinestatic |
|
inlinestatic |