[
https://issues.apache.org/jira/browse/ARROW-16423?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17530926#comment-17530926
]
Will Jones commented on ARROW-16423:
------------------------------------
Thanks.
I tried to reproduce again on 7.0.0, and got a segfault that wasn't solved by
the above snippet. It also doesn't fail on the same iteration each time, so
seems more like what you describe. I am not able to reproduce the same issue on
the dev edition, which means it will likely be fixed in the upcoming 8.0.0
release. I'll look a little bit more to check.
> [R] arrow/dplyr: simple join and collect crashes session
> --------------------------------------------------------
>
> Key: ARROW-16423
> URL: https://issues.apache.org/jira/browse/ARROW-16423
> Project: Apache Arrow
> Issue Type: Bug
> Components: R
> Affects Versions: 7.0.0
> Reporter: Andrew C Thomas
> Priority: Minor
>
> Trying to do an inner join style filter on an open_dataset, and R crashes,
> but not reliably the first time. Sometimes takes a couple of tries until it
> does.
> Reprex follows.
> ------------------------------------------------------
> library (arrow)
> library (dplyr)
> library (tidyr)
> DataSet <- expand_grid (A = 1:10, B = 1:10, C = 1:10000) %>%
> group_by (A, B)
> write_dataset(DataSet, "TestBreakData")
> for (DoThisUntilItBreaks in 1:100) {
> message (DoThisUntilItBreaks)
> D2 <- open_dataset("TestBreakData") %>% inner_join (data.frame (A=1L,
> B=1:5)) %>% collect
> }
--
This message was sent by Atlassian Jira
(v8.20.7#820007)
