Demonstration on the polynomial Wigner Ville with the non-uniform FFT

last update: 2/15 (2018)



In [1]:

    
using PyPlot
import DSP



In [2]:

    
include("../juwvid.jl")









    Out[2]:





juwvid



In [3]:

    
# multi component data
nsample=512
t,x=sampledata.genmultifm623x(nsample);
PyPlot.plot(t,x)









    












    Out[3]:





1-element Array{PyCall.PyObject,1}:
 PyObject <matplotlib.lines.Line2D object at 0x7f6305fa3668>

polynomial WV with FFT



In [4]:

    
# the resolution is poor  
trfpo=polywv.tfrpowv(x,NaN,NaN,NaN,2,2);
a=juwplot.wtfrshow(abs.(trfpo),t[2]-t[1],t[1],t[end],NaN,NaN,0.7*3)
PyPlot.ylim(0,40)









    



Single S-method






    












    



Use fft.






    Out[4]:





(0, 40)



In [5]:

    
nnufft=1024
nsample=length(t)
js,je=juwutils.frequency_to_index([5.0,30.0], t[2]-t[1], nsample,nnufft)









    Out[5]:





2-element Array{Float64,1}:
  20.0391
 120.235

with NuFFT



In [6]:

    
# now, the resolution is fine  
fin=collect(linspace(js,je,nnufft));
trfpo=polywv.tfrpowv(x,NaN,NaN,fin,2,2);
a=juwplot.wtfrshow(abs.(trfpo),t[2]-t[1],t[1],t[end],NaN,NaN,0.7)









    



Single S-method






    












    



Use nufft.






    Out[6]:





PyObject <matplotlib.image.AxesImage object at 0x7f630519dd68>