commons-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Sébastien Brisard <>
Subject Re: [math] Major speed improvement to 1D FFT (MATH-732)
Date Tue, 07 Feb 2012 12:33:35 GMT
> I've been testing the new faster code in a real world application and I'm a big fan.
 But, I too would be hesitant to change the API.  There are several ways it could be changed:
> transform (double[] real_imag)
> or
> transform(double[] real, double[] imag)
> or
> transform(ComplexVector cvec)
> (where ComplexVector might internally be implemented as either a real and imaginary double
array, or the alternating real/imag array, or an array of Complex).
> Choosing among these and other alternatives (including the current) might best be left
to a post-3.0 discussion of optimal handling of Complex arrays in general.
> Bruce
OK, I realize this was a bad suggestion. I'll commit Kurt's patch
(including cosmetic changes), and that's all. I would be tempted to
expose the lower level fft method, though (the one which takes two
double[] as an argument). Would you allow me that?


To unsubscribe, e-mail:
For additional commands, e-mail:

View raw message