From: kk aKI on
I want to do following step but not sure how to do it

1. Take some irregularly sample spikes
2. Make it as a power of 2 to apply FFT later
3. Do convolution with Gaussian function or B spline
4. Now after convolution my sample are regular, now i am applying FFT
5. Some spectrum after FFT

Any body can give me idea how to do this in matlab, when i do convolution with Gaussian or with B spline people call this gridding.......It is regularising the whole irregular sampled signal...

I want to see following figures to understand it better

1. Showing irregular data
2. showing convolution between signal and function
3. showing regularised data


Gaussian used in paper is exp(-b*x.^2); .....any help will be appreciated.....I am just too curious to try out this algorithm.....but written few steps not working for me,,,,,



Thanks

KK
From: TideMan on
On Feb 11, 1:26 pm, "kk aKI" <akikumar1...(a)gmail.com> wrote:
> I want to do following step but not sure how to do it
>
> 1. Take some irregularly sample spikes
> 2. Make it as a power of 2 to apply FFT later
> 3. Do convolution with Gaussian function or B spline
> 4. Now after convolution my sample are regular, now i am applying FFT
> 5. Some spectrum after FFT
>
> Any body can give me idea how to do this in matlab, when i do convolution with Gaussian or with B spline people call this gridding.......It is regularising the whole irregular sampled signal...
>
> I want to see following figures to understand it better
>
> 1. Showing irregular data
> 2. showing convolution between signal and function
> 3. showing regularised data
>
> Gaussian used in paper is exp(-b*x.^2); .....any help will be appreciated......I am just too curious to try out this algorithm.....but written few steps not working  for me,,,,,
>
> Thanks
>
> KK

Well, come on, show us the few steps you've written that are not
working.
We prefer to correct your code than to compose new code for you.
Who knows, you might be a homework cheat, wanting us to do your
assignment for you.
From: kk aKI on
TideMan <mulgor(a)gmail.com> wrote in message <a9763995-8c61-454b-96d0-576bd1077a55(a)k36g2000prb.googlegroups.com>...
> On Feb 11, 1:26 pm, "kk aKI" <akikumar1...(a)gmail.com> wrote:
> > I want to do following step but not sure how to do it
> >
> > 1. Take some irregularly sample spikes
> > 2. Make it as a power of 2 to apply FFT later
> > 3. Do convolution with Gaussian function or B spline
> > 4. Now after convolution my sample are regular, now i am applying FFT
> > 5. Some spectrum after FFT
> >
> > Any body can give me idea how to do this in matlab, when i do convolution with Gaussian or with B spline people call this gridding.......It is regularising the whole irregular sampled signal...
> >
> > I want to see following figures to understand it better
> >
> > 1. Showing irregular data
> > 2. showing convolution between signal and function
> > 3. showing regularised data
> >
> > Gaussian used in paper is exp(-b*x.^2); .....any help will be appreciated.....I am just too curious to try out this algorithm.....but written few steps not working  for me,,,,,
> >
> > Thanks
> >
> > KK
>
> Well, come on, show us the few steps you've written that are not
> working.
> We prefer to correct your code than to compose new code for you.
> Who knows, you might be a homework cheat, wanting us to do your
> assignment for you.

Well i dont think my code is usable enough or correct enough but i will be posting my code soon...with some questions

I found the same type of algorithm here
http://www.mathworks.com/matlabcentral/fileexchange/25135-nufft-nfft-usfft
But its based on greengard paper, what i am looking for is

http://scitation.aip.org/getabs/servlet/GetabsServlet?prog=normal&id=GPYSA7000064000002000539000001&idtype=cvips&gifs=yes&ref=no

I need some time to reproduce this, But sure i will

Thanks

Aki