Map reduce example to classify cluster and values

cluster_count.py

__author__ = 'uolter'

import map_reduce

def mapper(input_key, input_value):

    def cut_and_clean_value(cluster):
        """
        :param cluster:  string in the format <cluster>:<value>
        :return: touple cluster and value. If value is NaN return 0
        """

        ret = cluster.split(":")

        if ret[1] == 'NaN':
            ret[1] = 0.0
        val = round(float(ret[1]), 2)

        if val >= 0.80:
            val = 4
        elif val >= 0.60:
            val = 3
        elif val >= 0.40:
            val = 2
        elif val >= 0.20:
            val = 1
        else:
            val = 0

        return ret[0], val

    return [cut_and_clean_value(c) for c in input_value.split(";")]


def reducer(intermediate_key, intermediate_value_list):
    return intermediate_key, sum(intermediate_value_list)

def main():
    input_file = 'text/input.txt'
    i = {}
    with open(input_file) as file:
        data = file.readlines()
        for d in data:
            user_id = d[:9]
            i[user_id] = d[10:]

        print map_reduce.map_reduce(i, mapper, reducer)


if __name__ == '__main__':
    main()