Help protect the Great Barrier Reef with TensorFlow on Kaggle Join Challenge

tf.experimental.numpy.float16

Half-precision floating-point number type.

Inherits From: inexact

:Character code: 'e' :Canonical name: numpy.half :Alias on this platform (Linux x86_64): numpy.float16: 16-bit-precision floating-point number type: sign bit, 5 bits exponent, 10 bits mantissa.

T Scalar attribute identical to the corresponding array attribute.

Please see ndarray.T.

base Scalar attribute identical to the corresponding array attribute.

Please see ndarray.base.

data Pointer to start of data.
dtype Get array data-descriptor.
flags The integer value of flags.
flat A 1-D view of the scalar.
imag The imaginary part of the scalar.
itemsize The length of one element in bytes.
nbytes The length of the scalar in bytes.
ndim The number of array dimensions.
real The real part of the scalar.
shape Tuple of array dimensions.
size The number of elements in the gentype.
strides Tuple of bytes steps in each dimension.

Methods

all

Scalar method identical to the corresponding array attribute.

Please see ndarray.all.

any

Scalar method identical to the corresponding array attribute.

Please see ndarray.any.

argmax

Scalar method identical to the corresponding array attribute.

Please see ndarray.argmax.

argmin

Scalar method identical to the corresponding array attribute.

Please see ndarray.argmin.

argsort

Scalar method identical to the corresponding array attribute.

Please see ndarray.argsort.

as_integer_ratio

half.as_integer_ratio() -> (int, int)

Return a pair of integers, whose ratio is exactly equal to the original floating point number, and with a positive denominator. Raise OverflowError on infinities and a ValueError on NaNs.

np.half(10.0).as_integer_ratio()
(10, 1)
np.half(0.0).as_integer_ratio()
(0, 1)
np.half(-.25).as_integer_ratio()
(-1, 4)

astype

Scalar method identical to the corresponding array attribute.</