Irregular fourier matrix
in [Matlab]
|
From: kk KKsingh on 12 Jul 2010 16:07 I am doing some thing wrong here ! I am only concern with the spectrum... I have 750 samples My sampling rate is 4.00e^-10; f=-N/2:N/2-1 I am applying simple NDFT or Greg DFTGH ( As these lines are from greg code) on it Which is some thing like this....This is irregular signal not a regular one so taking real values in place of t instead of set of integer x = x(:); % Format 'x' into a column vector t = t(:); % Format 't' into a column vector f = f(:); % Format 'f' into a column vector dt = diff(t); % asymmetric "dt" dts = 0.5*([dt; 0]+[0; dt]); % symmetric "dt" meanx = sum(x.*dts)/sum(dts); N=length(x); df = diff(f); % asymmetric "df" dfs = 0.5*([df; 0]+[0; df]); % symmetric "df" W = exp(-2*pi*j * f*t'); XFT = W * (x.*dts); freqaxis=1/dt/N*[f] (Assuming i have even number of samples) plot(freqaxis,abs(XFT)) But I am getting weird output ! is some thing is wrong with the plotting..or freqaxias........i am saying weird because it not looking like spectrum not even close to FFT thing
From: kk KKsingh on 12 Jul 2010 18:34 "kk KKsingh" <akikumar1983(a)gmail.com> wrote in message <i1fsla$3dh$1(a)fred.mathworks.com>... > I am doing some thing wrong here ! I am only concern with the spectrum... > > I have 750 samples > My sampling rate is 4.00e^-10; > f=-N/2:N/2-1 > I am applying simple NDFT or Greg DFTGH ( As these lines are from greg code) on it Which is some thing like this....This is irregular signal not a regular one so taking real values in place of t instead of set of integer > > x = x(:); % Format 'x' into a column vector > t = t(:); % Format 't' into a column vector > f = f(:); % Format 'f' into a column vector > dt = diff(t); % asymmetric "dt" > dts = 0.5*([dt; 0]+[0; dt]); % symmetric "dt" > meanx = sum(x.*dts)/sum(dts); > N=length(x); > df = diff(f); % asymmetric "df" > dfs = 0.5*([df; 0]+[0; df]); % symmetric "df" > > W = exp(-2*pi*j * f*t'); > > XFT = W * (x.*dts); > > freqaxis=1/dt/N*[f] (Assuming i have even number of samples) > > plot(freqaxis,abs(XFT)) > > But I am getting weird output ! is some thing is wrong with the plotting..or freqaxias........i am saying weird because it not looking like spectrum not even close to FFT thing Using this my X axis is same as in FFT, but in y axis amplitude is 10^-4 in NDFT and 10^14 in FFT ...Can any one tell me what is wrong
From: Greg Heath on 13 Jul 2010 09:23 On Jul 12, 4:07 pm, "kk KKsingh" <akikumar1...(a)gmail.com> wrote: > I am doing some thing wrong here ! I am only concern with the spectrum... > > I have 750 samples > My sampling rate is 4.00e^-10; N = 750 Fs = 4e-10 % THAT HAS GOT TO BE A MISTAKE! df = Fs/N > f=-N/2:N/2-1 f = df*(-N/2:N/2-1); > I am applying simple NDFT or Greg DFTGH ( As these lines are from greg code) on it Which is some thing like this....This is irregular signal not a regular one so taking real values in place of t instead of set of integer > > x = x(:); % Format 'x' into a column vector > t = t(:); % Format 't' into a column vector > f = f(:); % Format 'f' into a column vector > dt = diff(t); % asymmetric "dt" > dts = 0.5*([dt; 0]+[0; dt]); % symmetric "dt" > meanx = sum(x.*dts)/sum(dts); > N=length(x); > df = diff(f); % asymmetric "df" > dfs = 0.5*([df; 0]+[0; df]); % symmetric "df" > > W = exp(-2*pi*j * f*t'); > > XFT = W * (x.*dts); > > freqaxis=1/dt/N*[f] (Assuming i have even number of samples) In correct. dt is a vector (see above) > plot(freqaxis,abs(XFT)) plot(f,abs(XFT)) > But I am getting weird output ! is some thing is wrong with the plotting...or freqaxias........i am saying weird because it not looking like spectrum not even close to FFT thing Thats the result of incorrectly using 1/dt/N instead of (Fs/N). Hope this helps. Greg
From: Greg Heath on 13 Jul 2010 09:28 On Jul 12, 6:34 pm, "kk KKsingh" <akikumar1...(a)gmail.com> wrote: > "kk KKsingh" <akikumar1...(a)gmail.com> wrote in message <i1fsla$3d...(a)fred..mathworks.com>... > > I am doing some thing wrong here ! I am only concern with the spectrum.... > > > I have 750 samples > > My sampling rate is 4.00e^-10; > > f=-N/2:N/2-1 > > I am applying simple NDFT or Greg DFTGH ( As these lines are from greg code) on it Which is some thing like this....This is irregular signal not a regular one so taking real values in place of t instead of set of integer > > > x = x(:); % Format 'x' into a column vector > > t = t(:); % Format 't' into a column vector > > f = f(:); % Format 'f' into a column vector > > dt = diff(t); % asymmetric "dt" > > dts = 0.5*([dt; 0]+[0; dt]); % symmetric "dt" > > meanx = sum(x.*dts)/sum(dts); > > N=length(x); > > df = diff(f); % asymmetric "df" > > dfs = 0.5*([df; 0]+[0; df]); % symmetric "df" > > > W = exp(-2*pi*j * f*t'); > > > XFT = W * (x.*dts); > > > freqaxis=1/dt/N*[f] (Assuming i have even number of samples) > > > plot(freqaxis,abs(XFT)) > > > But I am getting weird output ! is some thing is wrong with the plotting..or freqaxias........i am saying weird because it not looking like spectrum not even close to FFT thing > > Using this my X axis is same as in FFT, but in y axis amplitude is 10^-4 in NDFT and 10^14 in FFT ...Can any one tell me what is wrong- DFT includes dt. Therefore multiply XFFT by mean(dt) ~ 1/Fs = 0.25 * 10^10. Hope this helps. Greg
|
Pages: 1 Prev: Matrix indexing with two arrays question Next: Faster way to draw a solid circle |