Skip to contents

Computes the priority weight vector \(W\) and exact Lagrange multiplier \(\lambda\) from a Pairwise Reciprocal Matrix (PRM) A using the augmented Lagrangian Inverse Gram Matrix (LIGM) closed-form KKT system solution \([W, \lambda] = q \bar{G}^{-1}\) (Theorem 4; Yuen, 2024), eliminating the requirement for post-hoc vector normalization.

Usage

lagrangeIGM(A, r = 0, detail = FALSE)

LIGM(A, r = 0, detail = FALSE)

Arguments

A

Numeric square matrix of dimension n x n representing a Pairwise Reciprocal Matrix (PRM).

r

Numeric scalar parameter for the Gram matrix shift (default: 0).

detail

Logical. If TRUE, returns a list containing intermediate matrices (W, G, G_inv) (default: FALSE).

Value

If detail = FALSE, a numeric vector of length n + 1 containing the priority weights \(W\) and the Lagrange multiplier \(\lambda\). If detail = TRUE, a list containing intermediate calculation steps and matrices.

References

Yuen, K. K. F. (2024). Inverse Gram Matrix Methods for Prioritization in Analytic Hierarchy Process: Explainability of Weighted Least Squares Optimization Method. arXiv preprint arXiv:2401.01190. https://doi.org/10.48550/arXiv.2401.01190

Author

Kevin Kam Fung Yuen

Examples

# Define a 3x3 Pairwise Reciprocal Matrix
A <- matrix(c(1, 2, 5,
              1/2, 1, 3,
              1/5, 1/3, 1), nrow = 3, byrow = TRUE)

# Compute priority weights via lagrangeIGM (r = 0)
W <- lagrangeIGM(A, r = 0)
print(W)
#> [1] 0.5858925 0.3014962 0.1126112 0.9976250