Xiangrui Meng created SPARK-10329:
-------------------------------------
Summary: Cost RDD in k-means initialization is not
storage-efficient
Key: SPARK-10329
URL: https://issues.apache.org/jira/browse/SPARK-10329
Project: Spark
Issue Type: Improvement
Components: MLlib
Affects Versions: 1.4.1, 1.3.1, 1.5.0
Reporter: Xiangrui Meng
Currently we use `RDD[Vector]` to store point cost during k-means||
initialization, where each `Vector` has size `runs`. This is not
storage-efficient because `runs` is usually 1 and then each record is a Vector
of size 1. What we need is just the 8 bytes to store the cost, but we introduce
two objects (DenseVector and its values array), which could cost 16 bytes. That
is 200% overhead. Thanks [~Grace Huang] and Jiayin Hu from Intel for reporting
this issue!
There are several solutions:
1. Use `RDD[Array[Double]]` instead of `RDD[Vector]`, which saves 8 bytes per
record.
2. Use `RDD[Array[Double]]`) but batch the values for storage, e.g. each
`Array[Double]` object covers 1024 instances, which could remove most of the
overhead.
Besides, using MEMORY_AND_DISK instead of MEMORY_ONLY could prevent cost RDDs
kicking out the training dataset from memory.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
