[
https://issues.apache.org/jira/browse/SPARK-19771?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15893508#comment-15893508
]
Yun Ni commented on SPARK-19771:
[~merlin] What you are suggesting is to hash each AND hash vector into a
[
https://issues.apache.org/jira/browse/SPARK-19771?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15893117#comment-15893117
]
Mingjie Tang commented on SPARK-19771:
--
(1) because you need to explode each tuple. For example
[
https://issues.apache.org/jira/browse/SPARK-19771?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15893097#comment-15893097
]
Yun Ni commented on SPARK-19771:
[~merlin]
(1) The computation cost is NumHashFunctions because we go
[
https://issues.apache.org/jira/browse/SPARK-19771?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15889180#comment-15889180
]
Mingjie Tang commented on SPARK-19771:
--
If we follow the AND-OR framework, one optimization is here:
