diff options
author | Davies Liu <davies.liu@gmail.com> | 2014-08-24 21:16:05 -0700 |
---|---|---|
committer | Josh Rosen <joshrosen@apache.org> | 2014-08-25 11:33:20 -0700 |
commit | b82da3d6924a5bd2139434ab05c2fd44914fda45 (patch) | |
tree | c798d0525d0c62b20ccd1a5f28a8ce77953a52a3 | |
parent | 749bddc85e76e0d1ded8d79058819335bd580741 (diff) | |
download | spark-b82da3d6924a5bd2139434ab05c2fd44914fda45.tar.gz spark-b82da3d6924a5bd2139434ab05c2fd44914fda45.tar.bz2 spark-b82da3d6924a5bd2139434ab05c2fd44914fda45.zip |
[SPARK-2871] [PySpark] add zipWithIndex() and zipWithUniqueId()
RDD.zipWithIndex()
Zips this RDD with its element indices.
The ordering is first based on the partition index and then the
ordering of items within each partition. So the first item in
the first partition gets index 0, and the last item in the last
partition receives the largest index.
This method needs to trigger a spark job when this RDD contains
more than one partitions.
>>> sc.parallelize(range(4), 2).zipWithIndex().collect()
[(0, 0), (1, 1), (2, 2), (3, 3)]
RDD.zipWithUniqueId()
Zips this RDD with generated unique Long ids.
Items in the kth partition will get ids k, n+k, 2*n+k, ..., where
n is the number of partitions. So there may exist gaps, but this
method won't trigger a spark job, which is different from
L{zipWithIndex}
>>> sc.parallelize(range(4), 2).zipWithUniqueId().collect()
[(0, 0), (2, 1), (1, 2), (3, 3)]
Author: Davies Liu <davies.liu@gmail.com>
Closes #2092 from davies/zipWith and squashes the following commits:
cebe5bf [Davies Liu] improve test cases, reverse the order of index
0d2a128 [Davies Liu] add zipWithIndex() and zipWithUniqueId()
(cherry picked from commit fb0db772421b6902b80137bf769db3b418ab2ccf)
Signed-off-by: Josh Rosen <joshrosen@apache.org>
-rw-r--r-- | python/pyspark/rdd.py | 47 |
1 files changed, 47 insertions, 0 deletions
diff --git a/python/pyspark/rdd.py b/python/pyspark/rdd.py index 3eefc878d2..f0706d846d 100644 --- a/python/pyspark/rdd.py +++ b/python/pyspark/rdd.py @@ -1715,6 +1715,53 @@ class RDD(object): other._jrdd_deserializer) return RDD(pairRDD, self.ctx, deserializer) + def zipWithIndex(self): + """ + Zips this RDD with its element indices. + + The ordering is first based on the partition index and then the + ordering of items within each partition. So the first item in + the first partition gets index 0, and the last item in the last + partition receives the largest index. + + This method needs to trigger a spark job when this RDD contains + more than one partitions. + + >>> sc.parallelize(["a", "b", "c", "d"], 3).zipWithIndex().collect() + [('a', 0), ('b', 1), ('c', 2), ('d', 3)] + """ + starts = [0] + if self.getNumPartitions() > 1: + nums = self.mapPartitions(lambda it: [sum(1 for i in it)]).collect() + for i in range(len(nums) - 1): + starts.append(starts[-1] + nums[i]) + + def func(k, it): + for i, v in enumerate(it, starts[k]): + yield v, i + + return self.mapPartitionsWithIndex(func) + + def zipWithUniqueId(self): + """ + Zips this RDD with generated unique Long ids. + + Items in the kth partition will get ids k, n+k, 2*n+k, ..., where + n is the number of partitions. So there may exist gaps, but this + method won't trigger a spark job, which is different from + L{zipWithIndex} + + >>> sc.parallelize(["a", "b", "c", "d", "e"], 3).zipWithUniqueId().collect() + [('a', 0), ('b', 1), ('c', 4), ('d', 2), ('e', 5)] + """ + n = self.getNumPartitions() + + def func(k, it): + for i, v in enumerate(it): + yield v, i * n + k + + return self.mapPartitionsWithIndex(func) + def name(self): """ Return the name of this RDD. |