[
https://issues.apache.org/jira/browse/SPARK-4148?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Josh Rosen updated SPARK-4148:
------------------------------
Labels: (was: backport-needed)
> PySpark's sample uses the same seed for all partitions
> ------------------------------------------------------
>
> Key: SPARK-4148
> URL: https://issues.apache.org/jira/browse/SPARK-4148
> Project: Spark
> Issue Type: Bug
> Components: PySpark
> Affects Versions: 1.0.2, 1.1.0
> Reporter: Xiangrui Meng
> Assignee: Xiangrui Meng
> Fix For: 1.1.1, 1.2.0, 1.0.3
>
>
> The current way of seed distribution makes the random sequences from
> partition i and i+1 offset by 1.
> {code}
> In [14]: import random
> In [15]: r1 = random.Random(10)
> In [16]: r1.randint(0, 1)
> Out[16]: 1
> In [17]: r1.random()
> Out[17]: 0.4288890546751146
> In [18]: r1.random()
> Out[18]: 0.5780913011344704
> In [19]: r2 = random.Random(10)
> In [20]: r2.randint(0, 1)
> Out[20]: 1
> In [21]: r2.randint(0, 1)
> Out[21]: 0
> In [22]: r2.random()
> Out[22]: 0.5780913011344704
> {code}
> So the second value from partition 1 is the same as the first value from
> partition 2.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
