Fix parallel code for quanteda 4.0

[koheiw]

Update code for quanteda v4.0.

[kbenoit]

Any idea what's causing the segfaults from running this on Ubuntu? Is that because it needs quanteda 4.0?

> dfmat <- corpus_subset(data_corpus_inaugural, Year > 2000) %>%
+     tokens(remove_punct = TRUE) %>%
+     tokens_remove(stopwords("english")) %>%
+     dfm()
> (tstat1 <- textstat_simil(dfmat, method = "cosine", margin = "documents"))
 *** caught segfault ***

[koheiw]

It is probably because proxyC still uses RcppParallel. I have a branch to remove the dependency.

…

On Mon, Apr 1, 2024, 21:03 Kenneth Benoit ***@***.***> wrote: Any idea what's causing the segfaults from running this on Ubuntu? Is that because it needs *quanteda* 4.0? > dfmat <- corpus_subset(data_corpus_inaugural, Year > 2000) %>% + tokens(remove_punct = TRUE) %>% + tokens_remove(stopwords("english")) %>% + dfm() > (tstat1 <- textstat_simil(dfmat, method = "cosine", margin = "documents")) *** caught segfault *** — Reply to this email directly, view it on GitHub <#72 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABSEXIZ3MMSKCM7PJJHZS2TY3FEJLAVCNFSM6AAAAABFOMCO7OVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDAMRZGY2TGNBSGU> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[kbenoit]

Ok, sounds like then Resubmit proxyc Resubmit stats and models Then, quanteda All this because rcpparallel won’t maintain the tbb code consistently! Sent from Outlook for iOS<https://aka.ms/o0ukef>

________________________________ From: Kohei Watanabe ***@***.***> Sent: Monday, April 1, 2024 2:09:03 PM To: quanteda/quanteda.textstats ***@***.***> Cc: Benoit,KR ***@***.***>; Review requested ***@***.***> Subject: Re: [quanteda/quanteda.textstats] Fix parallel code for quanteda 4.0 (PR #72) It is probably because proxyC still uses RcppParallel. I have a branch to remove the dependency.

On Mon, Apr 1, 2024, 21:03 Kenneth Benoit ***@***.***> wrote: Any idea what's causing the segfaults from running this on Ubuntu? Is that because it needs *quanteda* 4.0? > dfmat <- corpus_subset(data_corpus_inaugural, Year > 2000) %>% + tokens(remove_punct = TRUE) %>% + tokens_remove(stopwords("english")) %>% + dfm() > (tstat1 <- textstat_simil(dfmat, method = "cosine", margin = "documents")) *** caught segfault *** — Reply to this email directly, view it on GitHub <#72 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABSEXIZ3MMSKCM7PJJHZS2TY3FEJLAVCNFSM6AAAAABFOMCO7OVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDAMRZGY2TGNBSGU> . You are receiving this because you authored the thread.Message ID: ***@***.***>

— Reply to this email directly, view it on GitHub<#72 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AAQUYZVUTTBERTI2FTJBY3LY3FL67AVCNFSM6AAAAABFOMCO7OVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDAMRZG4ZTGOJSHE>. You are receiving this because your review was requested.Message ID: ***@***.***>

[kbenoit]

@koheiw is this failing on the ubuntu platfoms because it is not getting the newest proxyC yet?