Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

MagnitudeCorrection trait should be public #274

Open
relf opened this issue Apr 7, 2021 · 0 comments · May be fixed by #275
Open

MagnitudeCorrection trait should be public #274

relf opened this issue Apr 7, 2021 · 0 comments · May be fixed by #275

Comments

@relf
Copy link

relf commented Apr 7, 2021

I tried to use TruncatedSvd with a generic Float type but it can not compile as it has to use the MagnitudeCorrection trait which is not exported from ndarray-linalg::lobpcg module. My test program here:

use ndarray::{arr1, arr2, Array1, Array2, ScalarOperand};
use ndarray_linalg::{close_l2, Lapack, TruncatedOrder, TruncatedSvd};  // should use MagnitudeCorrection as well
use num_traits::Float;

fn test_truncated_svd<F: Float + Lapack + ScalarOperand + Default>(  // should add MagnitudeCorrection trait bound
    a: Array2<F>,
) -> Array1<F> {
    let res = TruncatedSvd::new(a, TruncatedOrder::Largest)
        .precision(1e-5)
        .maxiter(10)
        .decompose(2)
        .unwrap();

    let (_, sigma, _) = res.values_vectors();
    sigma
}

fn main() {
    let a = arr2(&[[3., 2., 2.], [2., 3., -2.]]);

    let sigma = test_truncated_svd(a);

    close_l2(&sigma, &arr1(&[5.0, 3.0]), 1e-5);
}
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging a pull request may close this issue.

1 participant