Function random_correlation

pub fn random_correlation<F>(n: usize, seed: Option<u64>) -> Array2<F>
where
    F: Float + NumAssign + FromPrimitive + Clone + Debug + Sum + 'static,

Generate a random correlation matrix

A correlation matrix is symmetric positive semi-definite with ones on the diagonal. These matrices are commonly used in statistics, finance, and machine learning.

The method generates a random matrix, computes A^T * A, and then normalizes the result to have ones on the diagonal.

Arguments

• n - Dimension of the square correlation matrix
• seed - Optional seed for the random number generator

Returns

An nxn correlation matrix

Examples

use scirs2_linalg::random::random_correlation;

// Generate a 4x4 random correlation matrix
let c = random_correlation::<f64>(4, None);
assert_eq!(c.shape(), &[4, 4]);

// Check that diagonal elements are 1
for i in 0..4 {
    assert!((c[[i, i]] - 1.0).abs() < 1e-10);
}

// Check that off-diagonal elements are in [-1, 1]
for i in 0..4 {
    for j in (i+1)..4 {
        assert!(c[[i, j]] >= -1.0 && c[[i, j]] <= 1.0);
        // Also verify symmetry
        assert!((c[[i, j]] - c[[j, i]]).abs() < 1e-10);
    }
}