Distancia euclidiana y factor de correlación de Pearson. Similaridades en sets pequeños!

gistfile1.py

from math import sqrt

critics = {
    'Rod': {
        'Superman': 1.0,
        'Forest Gump': 7.0,
        'Transformers': 2.5,
    },
    'Novia': {
        'Superman': 0.8,
        'Forest Gump': 6.5,
        'Transformers': 2.3,
        'XXY': 1.0,
        'Munich': 3.0,
    },
    'Jose': {
        'Superman': 3.5,
        'Forest Gump': 3.0,
        'Transformers': 2.5,
        'XXY': 4.0,
        'Munich': 5.0,
    },
    'Andres': {
        'Superman': 1.0,
        'Forest Gump': 2.0,
        'Transformers': 7.0,
        'XXY': 1.0,
        'Munich': 4.5,
    }
}

# Returns a euclidian distance based similarity score for person 1 and person 2
def sim_distance(prefs, person1, person2):
    si=dict()
    for item in prefs[person1]:
        if item in prefs[person2]:
            si[item]=1


    if len(si) == 0:
        return 0

    sum_of_squares = sum(
        [pow(prefs[person1][item]-prefs[person2][item], 2) for item in si]
    )

    return 1/(1+sqrt(sum_of_squares))

# Returns the pearson correlation score for person 1 and person 2
def sim_pearson(prefs, person1, person2):
    si=dict()
    for item in prefs[person1]:
        if item in prefs[person2]:
            si[item]=1

    n=len(si)
    if n == 0:
        return 0

    # Add up all the prefs
    sum1 = sum([prefs[person1][item] for item in si])
    sum2 = sum([prefs[person2][item] for item in si])

    # Sum up the squares
    sum1Sq = sum([pow(prefs[person1][item], 2) for item in si])
    sum2Sq = sum([pow(prefs[person2][item], 2) for item in si])

    # Sum up the products
    pSum = sum([prefs[person1][item] * prefs[person2][item] for item in si])

    # Calculate pearson
    num = pSum-(sum1*sum2/n)
    den = sqrt(
        (sum1Sq-pow(sum1,2)/n) * (sum2Sq-pow(sum2,2)/n))
    if den == 0:
        return 0
    r = num/den

    return r

# Returns the best maches for person from the prefs dict.
# It calculates the similarity score for one user against all other users.
# It then sorts the scores in reverse order.

def topMatches(prefs, person, n=10, sim_algorithm=sim_pearson):
    scores = [
        (sim_algorithm(prefs, person, other), other)
        for other in prefs if other !=person
    ]

    # Sort the list
    scores.sort(reverse=True)
    return scores[0:n]

def getRecommendations(prefs, person, sim_algorithm=sim_pearson):
    totals=dict()
    simSums=dict()

    for user in prefs:
        if user == person:
            continue
        sim = sim_algorithm(prefs, person, user)
        if sim <= 0:
            continue

        for item in prefs[user]:
            # Only score for movies I haven't seen yet, which is pretty obvious
            if item not in prefs[person] or prefs[person][item] == 0:
                # Calculates similarity * score, so the closer a person is to me,
                # the score will be higher.
                totals.setdefault(item, 0)
                totals[item] += prefs[user][item] * sim

                # Sum of similarities
                simSums.setdefault(item, 0)
                simSums[item] += sim

    rankings = [
        (total/simSums[item], item) for item,total in totals.items()
    ]
    rankings.sort(reverse=True)

    return rankings