root / neighbors / neartag.py

import nearest

class NearestByTag(object):

    def __init__(self, tags):
        self.tags = tags
        self.db = nearest.DataSet(attrnames=tags, exclude=(-1,),)
        self.solver = nearest.NearestNeighborLearner()
        self.solver.train(self.db)

    def add_user(self, name, tags):
        values = []
        for tag in self.tags:
            if tag in tags:
                values.append('1')
            else:
                values.append('0')
        values.append(name)

        self.db.add_example(values)

    def neighbours(self, name, k=10):
        distances = []

        # find back the user
        current = None
        for user in self.db.examples:
            user_name = user[-1]
            if user_name == name:
                current = user

        if current is None:
            raise Exception('user %s not found' % name)

        for user in self.db.examples:
            user_name = user[-1]
            if user_name == name:
                continue
            distances.append((self.solver.distance(user, current), user_name))

        distances.sort()
        return distances[:k]