jax.numpy.linalg.eigvals#

jax.numpy.linalg.eigvals(a)[source]#

Compute the eigenvalues of a general matrix.

JAX implementation ofnumpy.linalg.eigvals().

Parameters:

a (ArrayLike) – array of shape(...,M,M) for which to compute the eigenvalues.

Returns:

An array of shape(...,M) containing the eigenvalues.

Return type:

Array

See also

• jax.numpy.linalg.eig(): computes eigenvalues eigenvectors of a general matrix.

• jax.numpy.linalg.eigh(): computes eigenvalues eigenvectors of a Hermitian matrix.

Notes

• This differs fromnumpy.linalg.eigvals() in that the return type ofjax.numpy.linalg.eigvals() is always complex64 for 32-bit input, andcomplex128 for 64-bit input.

• At present, non-symmetric eigendecomposition is only implemented on the CPU backend.

Examples

>>>a=jnp.array([[1.,2.],...[2.,1.]])>>>w=jnp.linalg.eigvals(a)>>>withjnp.printoptions(precision=2):...wArray([ 3.+0.j, -1.+0.j], dtype=complex64)