[SciPy-dev] New release on Monday
schofield at ftw.at
Sun Dec 11 15:55:06 CST 2005
On 11/12/2005, at 8:53 AM, Travis Oliphant wrote:
> I'm going to make a new release of scipy core on Monday evening (U.S.
> Mountain Time)
> unless I hear from anyone. I'll probably post new Windows binaries of
> full scipy
> at that point as well.
> Now's a good time for any bug-reports or outstanding issues.
Is there any chance of including type checking for this release for
unsafe in-place operations? Several people supported the idea in the
thread [In-place operators and casting] a few weeks ago. In summary,
the existing behaviour would still be achievable with an explicit cast:
>>> my_int_array += my_float_array.astype(int)
I was actually bitten by the current behaviour myself last week when
I was priming my maximum entropy module for the scipy tree. I spent
an hour trying to figure out why the computations were coming out
wrong. It turned out that the problem was that when converting my
home-made emptyarray() function to scipy_core's empty() I had
forgotten to specify doubles and my data were being silently
truncated. So I think type-checking wouldn't just help new users; it
would also save debugging time for users who know the existing
behaviour but, er, forget ;)
Well done with your recent optimizations!
More information about the Scipy-dev