[SciPy-dev] Problem with flapack
eric at scipy.org
Fri Feb 15 13:53:28 CST 2002
> thank you for fixing the problem. Here the result from a short test:
> >>> from scipy import *
> >>> from scipy import signal as s
> >>> s.poly(s.roots([1.,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20]))
> array([ 1.+0.00000000e+000j, 2.-1.89030962e-015j, 3.+3.97294109e-015j,
> 4.+2.16188134e-015j, 5.-5.21090017e-014j,
> 6.-1.44872571e-013j, 7.-2.18532601e-013j,
> 8.-2.36476441e-013j, 9.-2.38065611e-013j,
> 10.-4.36395620e-013j, 11.-9.58807728e-013j,
> 12.-1.45392938e-012j, 13.-1.52765658e-012j,
> 14.-1.21333755e-012j, 15.-6.78134691e-013j,
> 16.-2.67591593e-013j, 17.-7.35574790e-014j,
> 18.-8.47677514e-014j, 19.-1.43588852e-013j,
> It works fine. It should be possible to polish the values, in order to get
> rid of the small imaginary error, in the function poly.
Yeah, I noticed that too. I thought we already had done this??? Didn't Travis
O. put this in at one point. Maybe it was on a different set of functions.
Anyway, this should be added...
More information about the Scipy-dev