Gaussian line profile using numeric python

[Robert Amesz]

Konrad Hinsen wrote:

>"John J. Lee" <phrxy at csv.warwick.ac.uk> writes:
>
>> > 2) If you work with large data sets, you ought to use FFTs to
>> >    compute the convolution, that's O(N*log(N)) in the size of the
>> >    data set, instead of O(N**2) for the straightforward method.
>> >    Any good book on FFTs should explain how this works in detail.
>> 
>> but you probably don't need to know in detail, it boils down to
>> 
>> inv_FFT(FFT(data)*FFT(kernel))
>
>Except for zero-padding in case of non-periodic signals. That's why
>I always recommend to look at a detailed description before doing
>this; some of my colleagues keep telling me that FFT techniques "don't
>work", and I suspect this is the reason.

Well, why dont you/they try a different kind of integral transform, 
that might give a better result. Perhaps something like wavelets? Of 
course, a fast algorithm must be available to do the transform and its 
inverse but if there is it might be worth to do some experimenting.


Robert Amesz