Calculate the average of a dataset.

spark_average.py

    def average(self,dataset):
        output = dataset.map(lambda x:(x[1],x[2] ))
        output = output.groupByKey()
        output = output.map(lambda x:(x[0], len(x[1]),x[1]))
        output = output.map(lambda b:(b[0], sum(b[2])/b[1]))
        return output