[SciPy-user] Bad publicity, was:Re: problems with signal-functions impulse and step ...

Stef Mientki s.mientki@ru...
Mon Apr 16 13:46:44 CDT 2007

Stefan van der Walt wrote:
> On Mon, Apr 16, 2007 at 10:01:14AM +0200, Stef Mientki wrote:
>>> The function signature for signal.impulse is
>>> signal.impulse(system, X0=None, T=None, N=None)
>>> Where
>>>       system -- an instance of the LTI class or a tuple with 2, 3, or 4
>>>                 elements representing (num, den), (zero, pole, gain), or
>>>                 (A, B, C, D) representation of the system.
>>> (By the way, IPython is extremely useful in investigating this sort of
>>> thing -- you simply type signal.impulse?<enter>)
>>> So, I think what you want to do is:
>>> aa,bb = signal.impulse(filt_1[:2])
>> thanks Stéfan,
>> but it doesn't seem to work either.
hello Stéfan (and others),
> Are you saying that the following code snippet crashes?
> b,a = signal.iirdesign(0.06,0.006,1,50,0,'butter')
> signal.impulse((b,a))
yes (to be sure, I copied them again from your email ;-)
> If so, which version of scipy are you using?  This works fine on 0.5.3
> r2897.
Python 2.4.3 - Enthought Edition 1.0.0 (#69, Aug  2 2006, 12:09:59) [MSC 
v.1310 32 bit (Intel)] on win32.
Sorry, probably you mean a version of some sublibrary,
I don't know which library you mean and
I don't know how I can get that version.

Of course this is not very important to me,
because I can get the step and impuls response just as easy from lfilter.

But I think this is (very) bad publicity for SciPy,
after all, everyone starting with (deterministic) signal analysis,
will start with some simple filters,
and the first thing they want to see is amplitude response,
the second thing the want to see is the impuls response.
If indeed these functions are so buggy,
it would be better to just remove them,
because every beginner knows how to determine the impuls response, by 
filtering a Dirac puls.

just my 2 cents ;-)
But anyway thanks for the answers Stéfan !

Stef Mientki

More information about the SciPy-user mailing list