On 10/11/06, Greg Willden <[EMAIL PROTECTED]> wrote:
Perhaps you could explain *why* you want the schoolbook round? Given that floating point is inherently inaccurate you would have to expect to produce a lot of numbers exactly of the form x.5 *without errors*, which means you probably don't need round to deal with it. Anyway, absent a flag somewhere, you can do something like (x + sign(x)*.5).astype(int).
Chuck
Hi All,
I've read discussions in the archives about how round() "rounds to even" and how that is supposedly better.
But what I haven't been able to find is "What do I use if I want the regular old round that you learn in school?"
Perhaps you could explain *why* you want the schoolbook round? Given that floating point is inherently inaccurate you would have to expect to produce a lot of numbers exactly of the form x.5 *without errors*, which means you probably don't need round to deal with it. Anyway, absent a flag somewhere, you can do something like (x + sign(x)*.5).astype(int).
Chuck
------------------------------------------------------------------------- Using Tomcat but need to do more? Need to support web services, security? Get stuff done quickly with pre-integrated technology to make your job easier Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
_______________________________________________ Numpy-discussion mailing list Numpy-discussion@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/numpy-discussion
