Pyflix, confused about super() call

[process]

Anyone using Pyflix for the Netflix prize.

How can it call super to itself in its init-method?




---------------------


#!/usr/bin/env python

'''Sample baseline averaging algorithms.'''

import numpy as N
from pyflix.algorith ms import Algorithm


class MovieAverage(Al gorithm):
'''Baseline algorithm that computes the average of all the votes
for a movie
and predicts that for every user.

This algorithm returns an RMSE score of 1.0528 on the scrubbed
dataset.
'''

def __init__(self, training_set):
self._movie_ave rages = {}
super(MovieAver age,self).__ini t__(training_se t)

def __call__(self, movie_id, user_id):
try: return self._movie_ave rages[movie_id]
except KeyError:
avg =
N.average(self. _training_set.m ovie(movie_id). ratings())
self._movie_ave rages[movie_id] = avg
return avg


class UserAverage(Alg orithm):
'''Baseline algorithm that computes the average of all the votes
for a user
and predicts that for every movie.

This algorithm returns an RMSE score of 1.0688 on the scrubbed
dataset.
'''

def __init__(self, training_set):
self._user_aver ages = {}
super(UserAvera ge,self).__init __(training_set )

def __call__(self, movie_id, user_id):
try: return self._user_aver ages[user_id]
except KeyError:
avg =
N.average(self. _training_set.u ser(user_id).ra tings())
self._user_aver ages[user_id] = avg
return avg


class DoubleAverage(M ovieAverage,Use rAverage):
'''Returns the average of L{MovieAverage} and L{UserAverage}.

This algorithm returns an RMSE score of 1.0158 on the scrubbed
dataset.
'''

def __call__(self, movie_id, user_id):
return (MovieAverage._ _call__(self,mo vie_id,user_id) +
UserAverage.__c all__(self,movi e_id,user_id)) / 2

[Bruno Desthuilliers]

Re: Pyflix, confused about super() call

process a écrit :

Anyone using Pyflix for the Netflix prize.
>
How can it call super to itself in its init-method?
>

You mean :

class MovieAverage(Al gorithm):
def __init__(self, training_set):
self._movie_ave rages = {}

this line ?

super(MovieAver age,self).__ini t__(training_se t)


If that's what confuse you, I think you really should read the
FineManual(tm) instead of assuming Python is language X or Y.