[Numpy-discussion] Deprecate poly1d and replace with Poly1d ?

[Charles R Harris]

On Mon, Sep 28, 2009 at 2:59 PM, Robert Kern <robert.kern at gmail.com> wrote:

> On Mon, Sep 28, 2009 at 15:46, Charles R Harris
> <charlesr.harris at gmail.com> wrote:
>
> > The more basic problem here is making poly1d look like an array, which it
> > isn't. The array bit is an implementation detail and would be private in
> > C++. with an as_array method to retrieve the details if wanted.
>
> I'm pretty sure that it is an intentional public API and not an
> implementation detail. The __array__() method is not "making poly1d
> look like an array"; it is the standard name for such as_array()
> conversion methods.
>
>
Exactly, and that is why it is a design decision error. It *shouldn't* work
with as_array unless it is *an array*, which it isn't. Really

In [19]: sin(poly1d([1,2,3]))
Out[19]: array([ 0.84147098,  0.90929743,  0.14112001])

That makes no sense. On the other hand, it is difficult to make arrays of
poly1d, which does make sense because the polynomials are a commutative
ring.

Chuck
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/numpy-discussion/attachments/20090928/1875fc3c/attachment.html>