[SciPy-dev] r5268

Pauli Virtanen pav@iki...
Wed Dec 17 15:39:20 CST 2008


Wed, 17 Dec 2008 13:28:26 -0800, Jarrod Millman wrote:
[clip]
> Yes, the trunk is "frozen' until we branch for 0.70.  The only thing
> that should be committed to the trunk at this point is documentation
> improvements, new tests, or essential or trivial bug-fixes that have
> been discussed on the list.

Ok, I'll explain then also this: r5268

In the interp1d n-d cleanup I by mistake broke linear interpolation of 
complex values. This fixes it.

1D spline interpolation has never worked for complex-valued data, due to 
_fitpack.bisplev being written in C, so I changed interp1d to raise error 
in __init__ rather than later on in __call__.

Tests included, naturally.

-- 
Pauli Virtanen



More information about the Scipy-dev mailing list