Function opt_first_pivot 

pub fn opt_first_pivot<T, F>(
    f: &F,
    local_dims: &[usize],
    first_pivot: &MultiIndex,
    max_sweep: usize,
) -> MultiIndex
where
    T: Scalar + TTScalar,
    F: Fn(&MultiIndex) -> T,

Optimize the initial pivot by greedy coordinate-wise search.

Starting from first_pivot, sweeps through each dimension and replaces the current index with whichever value maximizes |f(pivot)|. Repeats until no improvement is found or max_sweep sweeps have been performed.

Arguments

• f – function to interpolate
• local_dims – number of values each index can take
• first_pivot – starting point for the search
• max_sweep – maximum number of full sweeps (1000 is a safe default)

Returns

The optimized pivot (multi-index with the largest |f| found).

Examples

use tensor4all_tensorci::opt_first_pivot;

let f = |idx: &Vec<usize>| (idx[0] as f64 + idx[1] as f64 + 1.0).powi(2);
let local_dims = vec![4, 4];
let start = vec![0, 0]; // f(0,0) = 1.0

let pivot = opt_first_pivot::<f64, _>(&f, &local_dims, &start, 1000);
// Should find the maximum: f(3,3) = 49.0
assert_eq!(pivot, vec![3, 3]);