org.tensorflow.op.linalg

Class SelfAdjointEig<T>

java.lang.Object

org.tensorflow.op.PrimitiveOp

org.tensorflow.op.linalg.SelfAdjointEig<T>

Type Parameters:

T - data type for e() output

All Implemented Interfaces:

Op

@Operator(group="linalg")
public final class SelfAdjointEig<T>
extends PrimitiveOp

Computes the eigen decomposition of one or more square self-adjoint matrices.

Computes the eigenvalues and (optionally) eigenvectors of each inner matrix in `input` such that `input[..., :, :] = v[..., :, :] * diag(e[..., :])`. The eigenvalues are sorted in non-decreasing order.

 # a is a tensor.
 # e is a tensor of eigenvalues.
 # v is a tensor of eigenvectors.
 e, v = self_adjoint_eig(a)
 e = self_adjoint_eig(a, compute_v=False)
 

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	SelfAdjointEig.Options Optional attributes for SelfAdjointEig

Field Summary

Fields inherited from class org.tensorflow.op.PrimitiveOp

operation

Method Summary

Modifier and Type	Method and Description
static SelfAdjointEig.Options	computeV(Boolean computeV)
static <T> SelfAdjointEig<T>	create(Scope scope, Operand<T> input, SelfAdjointEig.Options... options) Factory method to create a class wrapping a new SelfAdjointEig operation.
Output<T>	e() Eigenvalues.
Output<T>	v() Eigenvectors.

Methods inherited from class org.tensorflow.op.PrimitiveOp

equals, hashCode, op, toString

Methods inherited from class java.lang.Object

clone, finalize, getClass, notify, notifyAll, wait, wait, wait

Method Detail

create

public static <T> SelfAdjointEig<T> create(Scope scope,
                                           Operand<T> input,
                                           SelfAdjointEig.Options... options)

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

Parameters:

scope - current scope

input - `Tensor` input of shape `[N, N]`.

options - carries optional attributes values

Returns:

a new instance of SelfAdjointEig

computeV

public static SelfAdjointEig.Options computeV(Boolean computeV)

Parameters:

computeV - If `True` then eigenvectors will be computed and returned in `v`. Otherwise, only the eigenvalues will be computed.

e

public Output<T> e()

Eigenvalues. Shape is `[N]`.

v

public Output<T> v()

Eigenvectors. Shape is `[N, N]`.