ML Community Day is November 9! Join us for updates from TensorFlow, JAX, and more Learn more


LinearOperator acting like a [batch] square diagonal matrix.

Inherits From: LinearOperator, Module

Used in the notebooks

Used in the tutorials

This operator acts like a [batch] diagonal matrix A with shape [B1,...,Bb, N, N] for some b >= 0. The first b indices index a batch member. For every batch index (i1,...,ib), A[i1,...,ib, : :] is an N x N matrix. This matrix A is not materialized, but for purposes of broadcasting this shape will be relevant.

LinearOperatorDiag is initialized with a (batch) vector.

# Create a 2 x 2 diagonal linear operator.
diag = [1., -1.]
operator = LinearOperatorDiag(diag)

==> [[1.,  0.]
     [0., -1.]]

==> [2, 2]

==> scalar Tensor

x = ... Shape [2, 4] Tensor
==> Shape [2, 4] Tensor

# Create a [2, 3] batch of 4 x 4 linear operators.
diag = tf.random.normal(shape=[2, 3, 4])
operator = LinearOperatorDiag(diag)

# Create a shape [2, 1, 4, 2] vector.  Note that this shape is compatible
# since the batch dimensions, [2, 1], are broadcast to
# operator.batch_shape = [2, 3].
y = tf.random.normal(shape=[2, 1, 4, 2])
x = operator.solve(y)
==> operator.matmul(x) = y

Shape compatibility

This operator acts on [batch] matrix with compatible shape. x is a batch matrix with compatible shape for matmul and solve if

operator.shape = [B1,...,Bb] + [N, N],  with b >= 0
x.shape =   [C1,...,Cc] + [N, R],
and [C1,...,Cc] broadcasts with [B1,...,Bb] to [D1,...,Dd]


Suppose operator is a LinearOperatorDiag of shape [N, N], and x.shape = [N, R]. Then

  • operator.matmul(x) involves N * R multiplications.
  • operator.solve(x) involves N divisions and N * R multiplications.
  • operator.determinant() involves a size N reduce_prod.

If instead operator and x have shape [B1,...,Bb, N, N] and [B1,...,Bb, N, R], every operation increases in complexity by B1*...*Bb.

Matrix property hints

This LinearOperator is initialized with boolean flags of the form is_X, for X = non_singular, self_adjoint, positive_definite, square. These have the following meaning:

  • If is_X == True, callers should expect the operator to have the property X. This is a promise that should be fulfilled, but is not a runtime assert. For example, finite floating point precision may result in these promises being violated.
  • If is_X == False, callers should expect the operator to not have X.
  • If is_X == None (the default), callers should have no expectation either way.