Data Science. TileDB. Open Source. Quant Research. R. C++. Debian. Linux. Adjunct Clinical Professor, University of Illinois. Lots of coffee. And some running.
Asymptotically we will block everybody at which point Twitter will have become an email provider.
But they are currently so dumb and use multiple popular hash tags ("maximize eyeballs") that it really is like shooting fish in a barrel. Hence the 10min function, and high 'crit'.
No need for a Bayesian spam filter: the genius of the @csgillespie post was to just count hash tags.
What I need to add is a query about who is already blocked and then I have all for a cron job. Peace at last!
(#stats spam wave y'day, I used the function on 20+ senders).
I had placed a modified version of the @csgillespie function, with ample @rdatatable use, into a helper function in my `dang` package. This identifies the candidates, now I added a `post_silence()` function for mute + block + spam_report.
#rstats
A variant of the 'mute people with nc hashtags posts among n posts using term' function by @csgillespie is now in my `dang` package in a #tinyverse form requiring only #rdatatable.
To use it, make sure your Twitter token can read+write. Mine was still on read-only...
#rstats
Remember when @csgillespie showed an function for `rtweet` by @kearneymw to find hashtag spammers?
Tired of manually muting, blocking, spam reporting I added functions to
post_mute_nf (mute w/o follow)
post_block
post_spammer
which took ten minutes. Yay.
#rstats
Endorse -- just made this Dijon and Cognac Beef Stew as kiddo #2 goes back to her campus town in two days.
Takes a bit of time but fairly little effort, well worth it.
cooking.nytimes.com/recipes/…#NYTCooking
And @knapply_ and I have shown that `RcppSimdJson` can run quite a few performance circles around `jsonlite` (which remains wonderful and tested and complete but also quite a bit slower).
cran.r-project.org/package=R…
Thank you CRAN for smashing through another 1k barrier: now over 17k well-curated, always installable packages for us #rstats users. It's huge. And we appreciate it.
Longer and complete version of this rather fabulous live performance (running 6:31 rather than 2:20 below):
youtube.com/watch?v=fPkM8F0s…
Asking what the hell is going on never went out of style or scope.... So happy Hall of Fame induction birthday!
Could we have R 4.0.4 show a 404 error page, and release R 4.0.5 instead? Because interwebs?
Looking forward to a final 4.0.* release before the upcoming changes in 4.1.0 in three months.
#rstats
Big Thank You! to all who helped: @LukeTierney4 for initial issue and an excellent hint at fixing it, @kevin_ushey and others for comments---but chiefly @Enchufa2 for picking up my branch and running (much) further with it. Now in #Rcpp 1.0.6.3 after another PR polish. #rstats
This morning's new #rstats helper function: show all _compiled_ dependencies of a package (as Kurt, Uwe and I correctly suspected an Rcpp rev.dep. issue would go away under fresh builds).
Now in littler as new example 'compiledDeps.r'. Easy thanks to @rdatatable and `tools`.
How it started How it's going
I modified the intra-day SP500 tracker by @joshua_ulrich, the refactored code is in my `dang` package.
More changes to come incl a 24hr version for ES and alike (no free data though).
#rstats#rinfinance
Rcpp 1.0.6 on CRAN: Some updates
Seamless R and C++ Integration
dirk.eddelbuettel.com/blog/2…#rcpp#rstats
With special thanks to Walter Somerville, Mattias Ellert and Benjamin Christoffersen (all three twitter-less?) for contributed PRs.
/cc @kevin_ushey
Periodic reminder that @marutterstat is doing us #rstats users on @ubuntu a huge favour.
Relying on @rdatatable and my #Rcpp-based `RcppAPT` package, we see that we now have well over five thousand CRAN packages as pre-made binaries which is fabulous. Big thanks to Michael!