[SciPy-dev] reimplementation of lfilter

[Sturla Molden]

Sturla Molden skrev:
> It seems to be faster than signal.lfilter in most circumstances, 
> except one (annoyingly the most common). The most extreme case were 
> complex64 dtype with inplace filtering, for which I got 369% speed 
> improvement.
>
> Here are some timings (speed improvement over signal.lfilter in percent):

Sorry, my english were bad. The numbers are not speed "improvement", 
they are speed relative to signal.lfilter.

S.M.