tensor along dimension
axis using specified norm.
tensor, ord='euclidean', axis=None, name=None
tf.linalg.norm to compute the norm along
This function can compute several different vector norms (the 1-norm, the
Euclidean or 2-norm, the inf-norm, and in general the p-norm for p > 0) and
matrix norms (Frobenius, 1-norm, 2-norm and inf-norm).
Tensor of types
Order of the norm. Supported values are
np.inf and any positive real number yielding the corresponding
p-norm. Default is
'euclidean' which is equivalent to Frobenius norm if
tensor is a matrix and equivalent to 2-norm for vectors.
Some restrictions apply: a) The Frobenius norm
'fro' is not defined for
vectors, b) If axis is a 2-tuple (matrix norm), only
np.inf are supported. See the description of
on how to compute norms for a batch of vectors or matrices stored in a
None (the default), the input is considered a vector
and a single vector norm is computed over the entire set of values in the
norm(tensor, ord=ord) is equivalent to
norm(reshape(tensor, [-1]), ord=ord). If
axis is a Python integer, the
input is considered a batch of vectors, and
axis determines the axis in
tensor over which to compute vector norms. If
axis is a 2-tuple of
Python integers it is considered a batch of matrices and
the axes in
tensor over which to compute a matrix norm.
Negative indices are supported. Example: If you are passing a tensor that
can be either a matrix or a batch of matrices at runtime, pass
axis=[-2,-1] instead of
axis=None to make sure that matrix norms are
The name of the op.
Tensor with the same shape as
The computed norms with the same shape and dtype |
tensor but the
final axis is 1 instead. S