[Scipy-tickets] [SciPy] #1862: Allow canonical keyword arguments for distribution parameters

SciPy Trac scipy-tickets@scipy....
Sat Mar 9 10:45:05 CST 2013


#1862: Allow canonical keyword arguments for distribution parameters
-----------------------------+----------------------------------------------
 Reporter:  chrisfonnesbeck  |       Owner:  rgommers   
     Type:  enhancement      |      Status:  new        
 Priority:  low              |   Milestone:  Unscheduled
Component:  scipy.stats      |     Version:  0.11.0     
 Keywords:                   |  
-----------------------------+----------------------------------------------

Comment(by chrisfonnesbeck):

 I understand that this is harder to do than it would appear, but (at least
 from the viewpoint of someone who tries to teach statistics in Python) it
 would definitely add to clarity to be able to have named arguments. The
 signature binom(self, *args, **kwds) also hides the expected order of the
 parameters (n,p versus p,n).

 I think there *are* canonical parameter names for all the common
 distributions, in the sense that even though, for example, a beta
 distribution could be parameterized with alpha, beta or a, b, nobody would
 be confused as to which member of each pair was the scale or the shape
 parameter. Same for the normal distribution -- there is no ambiguity when
 the parameters are called mu and sigma.

 Of course, this is just a suggestion to help make scipy's distributions a
 little more user friendly.

-- 
Ticket URL: <http://projects.scipy.org/scipy/ticket/1862#comment:4>
SciPy <http://www.scipy.org>
SciPy is open-source software for mathematics, science, and engineering.


More information about the Scipy-tickets mailing list