RDD用法与实例（七）：collect和take

有时候collect结果会很奇怪。尽量用take，因为collect不是很能总能很好的收集数据。
如下这个例子中，collect就有了明显的不正确结果。

A = sc.parallelize(range(10))

x = 5
B = A.filter(lambda z: z < x)
# B.cache()
B.unpersist()
print(B.take(10))
print(B.collect())
x = 3
print(B.take(10))
print(B.collect())
# collect() doesn't always re-collect data - bad design!
# Always use take() instead of collect()

[0, 1, 2, 3, 4]
[0, 1, 2, 3, 4]
[0, 1, 2]
[0, 1, 2, 3, 4]