Spark - (Map|flatMap)

The map implementation in Spark of map reduce.

• map(func) returns a new distributed data set that's formed by passing each element of the source through a function.
• flatMap(func) similar to map but flatten a collection object to a sequence.

• PySpark: lambda function

rdd = sc.parallelize([1,2,3,4])
rdd.map(lambda x: x * 2).collect()

[2,4,6,8]

• def function

rdd = sc.parallelize([1,2,3,4])
def plus5(x):
    return x+5
rdd.map(plus5).collect()

[6, 7, 8, 9]

rdd = sc.parallelize([1,2,3])
rdd.flatMap(lambda x:[x,x+5]).collect()

[1,6,2,7,3,8]

sc.parallelize([[1,2],[3,4]]).flatMap(lambda x:x).collect()

[1, 2, 3, 4]

key value (tuple) transformation are supported

from operator import add
wordCounts = sc.parallelize([('rat', 2), ('elephant', 1), ('cat', 2)])
totalCount = (wordCounts
              .map(lambda (x,y): y)
              .reduce(add))
average = totalCount / float(len(wordCounts.collect()))
print totalCount
print round(average, 2)

5
1.67