tf.irfftnd  |  TensorFlow v2.16.1 (original) (raw)

ND inverse real fast Fourier transform.

View aliases

Compat aliases for migration

SeeMigration guide for more details.

tf.compat.v1.irfftnd

tf.irfftnd(
    input: Annotated[Any, TV_IRFFTND_Tcomplex],
    fft_length: Annotated[Any, _atypes.Int32],
    axes: Annotated[Any, _atypes.Int32],
    Treal: TV_IRFFTND_Treal = tf.dtypes.float32,
    name=None
) -> Annotated[Any, TV_IRFFTND_Treal]

Computes the n-dimensional inverse real discrete Fourier transform over designated dimensions of input. The designated dimensions of input are assumed to be the result of IRFFTND. The inner-most dimension contains thefft_length / 2 + 1 unique components of the DFT of a real-valued signal.

If fft_length[i]<shape(input)[i], the="" input="" is="" cropped.="" if="" fft_length[i]="">shape(input)[i], the input is padded with zeros. If fft_length is not given, the default shape(input) is used.</shape(input)[i],>

Axes mean the dimensions to perform the transform on. Default is to perform on all axes.

Args
input A Tensor. Must be one of the following types: complex64, complex128. A complex tensor.
fft_length A Tensor of type int32. An int32 tensor. The FFT length for each dimension.
axes A Tensor of type int32. An int32 tensor with a same shape as fft_length. Axes to perform the transform.
Treal An optional tf.DType from: tf.float32, tf.float64. Defaults to tf.float32.
name A name for the operation (optional).
Returns
A Tensor of type Treal.