Plan a solver for a problem.

Usage

plan_solver(problem, ...)

Arguments

problem: Eigencore eigen or SVD problem object.
...: Additional planning arguments passed to methods.

Value

A schema-versioned executable eigencore_plan containing the problem, requested size, original method descriptor, frozen route, canonical controls and planner policy, execution arguments, operator identity, serialization capability, and retained-memory metadata.

Details

The eigen and SVD methods accept their usual request and method arguments plus execution controls through .... Eigen plans freeze tol, maxit, vectors, certify, allow_dense_fallback, and an optional initial_subspace; SVD plans freeze tol, vectors, certify, and allow_dense_fallback. Use solve(plan) to execute those values or solve(plan, replan = TRUE) to make a fresh decision under current policy.

Examples

A <- diag(c(4, 3, 2, 1))
plan <- plan_solver(eigen_problem(A), k = 2)
plan$method
#> [1] "native dense Hermitian LAPACK fallback"
plan$reasons
#> [1] "structure: hermitian"                          
#> [2] "target: largest"                               
#> [3] "standard eigenproblem"                         
#> [4] "built-in dense operator has native block apply"