Svd

public final class Svd

Computes the eigen decomposition of a batch of self-adjoint matrices

(Note: Only real inputs are supported).

Computes the eigenvalues and eigenvectors of the innermost M-by-N matrices in tensor such that tensor[...,:,:] = u[..., :, :] * Diag(s[..., :]) * Transpose(v[...,:,:]).

Constants

String OP_NAME The name of this op, as known by TensorFlow core engine

Public Methods

static <T extends TType> Svd<T>
create(Scope scope, Operand<T> a, Long maxIter, Float epsilon, String precisionConfig)
Factory method to create a class wrapping a new Svd operation.
Output<T>
s()
Singular values.
Output<T>
u()
Left singular vectors.
Output<T>
v()
Right singular vectors.

Inherited Methods

Constants

public static final String OP_NAME

The name of this op, as known by TensorFlow core engine

Constant Value: "XlaSvd"

Public Methods

public static Svd<T> create (Scope scope, Operand<T> a, Long maxIter, Float epsilon, String precisionConfig)

Factory method to create a class wrapping a new Svd operation.

Parameters
scope current scope
a the input tensor.
maxIter maximum number of sweep update, i.e., the whole lower triangular part or upper triangular part based on parameter lower. Heuristically, it has been argued that approximately log(min (M, N)) sweeps are needed in practice (Ref: Golub & van Loan "Matrix Computation").
epsilon the tolerance ratio.
precisionConfig a serialized xla::PrecisionConfig proto.
Returns
  • a new instance of Svd

public Output<T> s ()

Singular values. The values are sorted in reverse order of magnitude, so s[..., 0] is the largest value, s[..., 1] is the second largest, etc.

public Output<T> u ()

Left singular vectors.

public Output<T> v ()

Right singular vectors.