[Numpy-discussion] ndarray sub-classing and append function

Hi, I am sub-classing numpy.ndarry for vector array representation. The append function is like this:     def append(self, other):        self = numpy.append(self, [other], axis=0) Example: vary = VectorArray([v1, v2]) #vary = numpy.append(vary, [v1], axis=0) vary.append(v1) The commented

Re: [Numpy-discussion] ndarray sub-classing and append function

It doesn't work because numpy.append(a, ...) doesn't modify the array a in-place: it returns a copy. Then in your append method, doing self = numpy.append(...) won't have any effect: in Python such a syntax means the self local variable will now point to the result of numpy.append, but it won't

[Numpy-discussion] Trac configuration tweak

Hi, I moved projects.scipy.org Tracs to run on mod_python (instead of CGI), in order to try to combat the present performance issues. Let's see if this helps with the database is locked problem. Please drop me a message if something stops working. Pauli

Re: [Numpy-discussion] YouTrack testbed

The idea is to allow people to test-out YouTrack for a few weeks and get to know it while we migrate bugs to it. it looks like it is straightforward to export the data out of YouTrack should we eventually decide to use something else. The idea is to host it on an external server (Rackspace

Re: [Numpy-discussion] Style for pad implementation in 'pad' namespace or functions under np.lib

1) The use of string constants to identify NumPy processes. It would seem better to use library defined constants (ufuncs?) for better future-proofing, maintenance, etc. I don't see how this would help with future-proofing or maintenance -- can you elaborate? If this were C, I'd agree;