Required for acquisitions in TbT rate that have switching harmonics
Now scipy.fft is used instead of numpy.fft since it is slightly faster (~20% in my PC. I tested in the control rooms PCs and the performance is the same for both packages).
Example considering the typical sizes of orbit data executed in my PC:
import numpy as np
import scipy as sp
Nsamps = 100_000
Nbpms = 160
mat = np.random.rand(Nsamps, Nbpms)
%timeit np_fft = np.fft.rfft(mat)
# result: 56.1 ms ± 107 µs per loop (mean ± std. dev. of 7 runs, 10 loops each)
%timeit sp_fft = sp.fft.rfft(mat)
# result: 44.5 ms ± 110 µs per loop (mean ± std. dev. of 7 runs, 10 loops each)
Required for acquisitions in TbT rate that have switching harmonics
Now scipy.fft is used instead of numpy.fft since it is slightly faster (~20% in my PC. I tested in the control rooms PCs and the performance is the same for both packages).
Example considering the typical sizes of orbit data executed in my PC: