randomPairs.py

#!/usr/bin/python3
import random
import math
from itertools import zip_longest
from pprint import pprint

def grouper(iterable, n, fillvalue=None):
    "Collect data into fixed-length chunks or blocks"
    # grouper('ABCDEFG', 3, 'x') --> ABC DEF Gxx"
    args = [iter(iterable)] * n
    return zip_longest(*args, fillvalue=fillvalue)

class OptionsHolder:
    def __init__(self):
        self.options = []
        self.weights = []
        
    def add_option(self, option):
        weight = 1
        for pair in option:
            r1 = pair[0][1]
            r2 = pair[1][1]
            addWeight = math.exp(-((r1 + r2)/2 - averageRating) ** 2 / D)
            #addWeight *= math.exp(-(r1 - r2) ** 2 / 2 / D)
            weight *= addWeight
        self.options.append(option)
        self.weights.append(weight)
        
    def select_option(self):
        #print(self.options)
        #print(self.weights)
        sumWeights = sum(self.weights)
        rnd = random.uniform(0, sumWeights)
        cur = 0
        for i in range(len(self.options)):
            cur += self.weights[i]
            #print(options[i], cur, weights[i])
            if cur > rnd:
                print("returning ", self.weights[i])
                return self.options[i]
       

#random.seed(124234)
data = open("users.txt").readlines()
data = [x.split() for x in data]
for x in data:
    x[1] = int(x[1])

averageRating = sum([x[1] for x in data]) / len(data)
D = sum(x[1]**2 for x in data)/len(data) - averageRating**2
print("average = ", averageRating, "  D = ", D, " ", math.sqrt(D))

D0 = D
D *= 0.5



first = OptionsHolder()

for i in range(100):
    random.shuffle(data)
    option = list(grouper(data, 2, ["_", averageRating - math.sqrt(D0)]))
    first.add_option(option)

option = first.select_option()
with open("first.txt", "w") as f:
    for pair in option:
        f.write(pair[0][0] + " " + str(pair[0][1]) + "   " + pair[1][0] + " " + str(pair[1][1]) + "\n")
        
firsts = []
seconds = []
for pair in option:
    firsts.append(pair[0])
    seconds.append(pair[1])

second = OptionsHolder()
for i in range(100):
    random.shuffle(seconds)
    option = list(zip(seconds, firsts))
    second.add_option(option)
    
option = second.select_option()
with open("second.txt", "w") as f:
    for pair in option:
        f.write(pair[0][0] + " " + str(pair[0][1]) + "   " + pair[1][0] + " " + str(pair[1][1]) + "\n")