On Nov 3, 10:41 pm, Felix Lawrence <fe...@physics.usyd.edu.au> wrote:
> Here's why there's a discrepancy:
> sage: import scipy.fftpack as Fourier
> sage: Fourier.ifft([95,20-25*i,-23,20+25*i])
> array([ 28.0+0.j, 29.5-0.j, 8.0+0.j, 29.5+0.j])
> sage: Fourier.ifft([95,20,-23,20])
> array([ 28.0+0.j, 29.5-0.j, 8.0+0.j, 29.5+0.j])
> sage: sage: Fourier.ifft([95,20-25j,-23,20+25j])
> array([ 28.+0.j, 42.+0.j, 8.+0.j, 17.+0.j])
>
> I encountered a similar bug in numpy last week when using
> numpy.loadtxt - any terms involving 'I' were ignored. I think it's an
> upstream problem. In my case I could get around it since loadtxt
> accepts a list of converters, and you send it this one:
> lambda s: complex(s.replace('i','j'))
I see - that makes sense, since numpy is a Python thing we don't need
an interface per se, and so it only accepts 'legitimate' Python
complex numbers. I'm just surprised it doesn't throw an error in that
case! I'm not sure upstream would say it's a problem - do you mind
asking the numpy list about this? I am not exactly a numerical guy,
and would probably ask the wrong question :)
> Does anyone know if this is fixed in newer versions of numpy/scipy?
I bet someone could test this if they've already built 4.6.1.alpha0 !
Volunteers?
- kcrisman
--
To post to this group, send email to sage-support@googlegroups.com
To unsubscribe from this group, send email to
sage-support+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/sage-support
URL: http://www.sagemath.org
