[
https://issues.apache.org/jira/browse/SPARK-6724?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14710278#comment-14710278
]
Feynman Liang commented on SPARK-6724:
--------------------------------------
An {{FPGrowth}} model just wraps the frequent itemsets and their corresponding
frequencies. However, we currently allow the items in the itemsets to be any
arbitrary class whereas allowing the model to be exported would at the very
least require serializable items.
PR 7320 actually makes a stronger assumption; the items must be primitive types
supported by {{spark.sql.types}}.
Is this assumption acceptable? If so, we should probably add some checks to
FPGrowth (and PrefixSpan since the frequent sequences implementation is very
similar) to enforce it.
> Model import/export for FPGrowth
> --------------------------------
>
> Key: SPARK-6724
> URL: https://issues.apache.org/jira/browse/SPARK-6724
> Project: Spark
> Issue Type: Sub-task
> Components: MLlib
> Affects Versions: 1.3.0
> Reporter: Joseph K. Bradley
> Priority: Minor
>
> Note: experimental model API
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
