mindscience.sciops.fft.ASD_RFFT

class mindscience.sciops.fft.ASD_RFFT[source]

1D real-to-complex FFT transform using Ascend NPU acceleration.

This operator performs 1D Real Fast Fourier Transform on real input tensors, optimized for Ascend NPU hardware acceleration.

Inputs:

  • xr (Tensor): Input real tensor with data type float32.

Outputs:

  • yr (Tensor): Real part of output complex tensor with data type float32.

  • yi (Tensor): Imaginary part of output complex tensor with data type float32.

Raises

ValueError – If input tensor data type is not float32.

Examples

>>> import mindspore as ms
>>> from mindscience.sciops.fft import ASD_RFFT
>>> xr = ms.Tensor([[1.0, 2.0, 3.0, 4.0]], ms.float32)
>>> asd_rfft = ASD_RFFT()
>>> yr, yi = asd_rfft(xr)
>>> print(yr.shape)
(1, 3)
>>> print(yi.shape)
(1, 3)