TensorFlow 1 version View source on GitHub

Computes the 1D Discrete Cosine Transform (DCT) of input.

    input, type=2, n=None, axis=-1, norm=None, name=None

Types I, II, III and IV are supported. Type I is implemented using a length 2N padded tf.signal.rfft. Type II is implemented using a length 2N padded tf.signal.rfft, as described here: Type 2 DCT using 2N FFT padded (Makhoul). Type III is a fairly straightforward inverse of Type II (i.e. using a length 2N padded tf.signal.irfft). Type IV is calculated through 2N length DCT2 of padded signal and picking the odd indices.


  • input: A [..., samples] float32/float64 Tensor containing the signals to take the DCT of.
  • type: The DCT type to perform. Must be 1, 2, 3 or 4.
  • n: The length of the transform. If length is less than sequence length, only the first n elements of the sequence are considered for the DCT. If n is greater than the sequence length, zeros are padded and then the DCT is computed as usual.
  • axis: For future expansion. The axis to compute the DCT along. Must be -1.
  • norm: The normalization to apply. None for no normalization or 'ortho' for orthonormal normalization.
  • name: An optional name for the operation.


A [..., samples] float32/float64 Tensor containing the DCT of input.


  • ValueError: If type is not 1, 2, 3 or 4, axis is not -1, n is not None or greater than 0, or norm is not None or 'ortho'.
  • ValueError: If type is 1 and norm is ortho.

Scipy Compatibility

Equivalent to scipy.fftpack.dct for Type-I, Type-II, Type-III and Type-IV DCT.