Prev: mexCallMatlab access violation
Next: measuring torque in simhydraulics
From: bovitch on 8 Jun 2010 18:27
HI
I have a matrix that I must calculate IFFT (log (Matrix(k, n))) for k = 0 ,...., K with K
the length of the DFT
please can you help me
From: Walter Roberson on 8 Jun 2010 18:33
bovitch wrote:

> I have a matrix that I must calculate IFFT (log (Matrix(k, n))) for k =
> 0 ,...., K with K
> the length of the DFT
> please can you help me

What does Matrix(k,n) represent? If it is subscripting, then you will have
difficulty because subscript 0 is not permitted for matrices. If you fix that,
then is n a vector? If not then you are subscripting a single element and the
ifft of a single element is not going to be very interesting.

Are you looking for this?

Results = zeros(K, length(n));
for k = 1 : K
Results(k, :) = ifft(log(Matrix(k, 1:10)));
end
From: bovitch on 8 Jun 2010 18:58
Walter Roberson <roberson(a)hushmail.com> wrote in message <humgid$jpn$1(a)canopus.cc.umanitoba.ca>...
> bovitch wrote:
>
> > I have a matrix that I must calculate IFFT (log (Matrix(k, n))) for k =
> > 0 ,...., K with K
> > the length of the DFT
> > please can you help me
>
> What does Matrix(k,n) represent? If it is subscripting, then you will have
> difficulty because subscript 0 is not permitted for matrices. If you fix that,
> then is n a vector? If not then you are subscripting a single element and the
> ifft of a single element is not going to be very interesting.
>
> Are you looking for this?
>
> Results = zeros(K, length(n));
> for k = 1 : K
> Results(k, :) = ifft(log(Matrix(k, 1:10)));
> end
the matrix is binary mask IBM , in order to transform the binary mask from t-f domain into the cepstral domain, i must calculate the inverse TFD{log (matrix(k,n)) for k=0,...,K} with K is the length of the DFT
From: Walter Roberson on 8 Jun 2010 19:20
bovitch wrote:
> Walter Roberson <roberson(a)hushmail.com> wrote in message
> <humgid$jpn$1(a)canopus.cc.umanitoba.ca>...
>> bovitch wrote:

>> > I have a matrix that I must calculate IFFT (log (Matrix(k, n))) for
>> k = > 0 ,...., K with K
>> > the length of the DFT
>> > please can you help me

>> What does Matrix(k,n) represent? If it is subscripting, then you will
>> have difficulty because subscript 0 is not permitted for matrices. If
>> you fix that, then is n a vector? If not then you are subscripting a
>> single element and the ifft of a single element is not going to be
>> very interesting.

> the matrix is binary mask IBM , in order to transform the binary mask
> from t-f domain into the cepstral domain, i must calculate the inverse
> TFD{log (matrix(k,n)) for k=0,...,K} with K is the length of the DFT

Well, you cannot do that, as it is not possible to index an array at element #0.

What size is Matrix? What size is n?

If Matrix is at least K+1 rows and n is a vector, then the code I provided
earlier should work (I think), except you might need to change the upper limit
of the 'for' loop to K+1 instead of K.

A question, though: when you write IFFT, are you implying the inverse of the
continuous fourier transform or the discrete fourier transform?
From: bovitch on 8 Jun 2010 19:45
Walter Roberson <roberson(a)hushmail.com> wrote in message <humjbs$nou$1(a)canopus.cc.umanitoba.ca>...
> bovitch wrote:
> > Walter Roberson <roberson(a)hushmail.com> wrote in message
> > <humgid$jpn$1(a)canopus.cc.umanitoba.ca>...
> >> bovitch wrote:
>
> >> > I have a matrix that I must calculate IFFT (log (Matrix(k, n))) for
> >> k = > 0 ,...., K with K
> >> > the length of the DFT
> >> > please can you help me
>
> >> What does Matrix(k,n) represent? If it is subscripting, then you will
> >> have difficulty because subscript 0 is not permitted for matrices. If
> >> you fix that, then is n a vector? If not then you are subscripting a
> >> single element and the ifft of a single element is not going to be
> >> very interesting.
>
> > the matrix is binary mask IBM , in order to transform the binary mask
> > from t-f domain into the cepstral domain, i must calculate the inverse
> > TFD{log (matrix(k,n)) for k=0,...,K} with K is the length of the DFT
>
> Well, you cannot do that, as it is not possible to index an array at element #0.
>
> What size is Matrix? What size is n?
>
> If Matrix is at least K+1 rows and n is a vector, then the code I provided
> earlier should work (I think), except you might need to change the upper limit
> of the 'for' loop to K+1 instead of K.
>
> A question, though: when you write IFFT, are you implying the inverse of the
> continuous fourier transform or the discrete fourier transform?
thanks
the size of matrix is 144*128
i use DFT (the discrete Fourier transform) but i dont know how implying the DFT of log(mask)
 |  Next  |  Last
Pages: 1 2
Prev: mexCallMatlab access violation
Next: measuring torque in simhydraulics