For our N-dimension solvers, we should look at maybe testing our solvers out with some of the test suites given here: https://github.com/devernay/cminpack/tree/master/examples and here: https://people.sc.fsu.edu/~jburkardt/f_src/test_nonlin/test_nonlin.html . These tests are from the MINPACK test suite for the nonlinear solvers and should at least provide confidence in any new N-dimension solvers that we might add. I only mention this as I've found that the Broyden test didn't catch an error in a newer solver I'm working on but one of the above tests did.
I can look at adding at least some of those with my up-coming PR for this new solver.
For our N-dimension solvers, we should look at maybe testing our solvers out with some of the test suites given here: https://github.com/devernay/cminpack/tree/master/examples and here: https://people.sc.fsu.edu/~jburkardt/f_src/test_nonlin/test_nonlin.html . These tests are from the MINPACK test suite for the nonlinear solvers and should at least provide confidence in any new N-dimension solvers that we might add. I only mention this as I've found that the Broyden test didn't catch an error in a newer solver I'm working on but one of the above tests did.
I can look at adding at least some of those with my up-coming PR for this new solver.