Hi Matthew
On Tue, Jan 02, 2007 at 02:30:34AM -0500, Matthew Koichi Grimes wrote:
> Is it hard to subclass recarrays? I'm currently working on a subclass
> that doesn't add much; just some methods that sets all the values to
> zero and other similarly trivial things. It'd be nice to make it work,
> but if I have to use external functions on a plain recarray instead of
> this subclass, I won't be found weeping in the corner.
Pierre wrote this useful wiki entry on subclassing:
http://www.scipy.org/Subclasses
It describes exactly what you want to do.
Find attached some example code.
Regards
Stéfan
import numpy as N
class SubRec(N.recarray):
dtype = N.dtype([('x','f8'),('dx','f8'),('deltas','f8')])
def __new__(subtype,shape):
data = N.zeros(shape,subtype.dtype)
return data.view(subtype)
def __array_finalize__(self,obj):
pass # for now
def clearGradient(self):
"""Fills all members other than x with zeros."""
self['dx'][:] = 0.
self['deltas'][:] = 0.
s = SubRec([3])
s['deltas'][:] = 1.
s['dx'][:] = 2.
print s
s.clearGradient()
print s
_______________________________________________
Numpy-discussion mailing list
Numpy-discussion@scipy.org
http://projects.scipy.org/mailman/listinfo/numpy-discussion
