Introduction - If you have any usage issues, please Google them yourself
The kernel matrix eigenvalue decomposition algorithm for DFT F (k, n) characteristic value And the feature vectors are constructed as the fractional powers of the DFT kernel matrix, which are used as Fa (k, n)
To calculate DFrFT.