Data Science. TileDB. Open Source. Quant Research. R. C++. Debian. Linux. Adjunct Clinical Professor, University of Illinois. Lots of coffee. And some running.
Also, updating 'leaf' (or 'near-leaf') packages with no (or few) reverse dependencies is still the best at CRAN: Seven minutes net. #ThankYouCRAN#RStats
It is of course also worth the occassional weekly wait (as for Rcpp), or test discussion. Quality matters, and CRAN rules.
Today's `r2u` example: installing `imputeTS` in one step, ten seconds (!!) for fifty-ish total packages, and without any BLAS/LAPACK hickups as the @StackOverflow question below experienced. Still rocks hard😀
stackoverflow.com/questions/…#rstats#r2u
Not so. My CRANberries code has been running (now on an hourly cronjob) for fifteen years. I have owned production jobs that ran for years.
CRAN ensures an absence of fundamental breakage. You can combine that with your knowledge and experience to pick packages that behave.
Achievement unlocked: A package of mine has been meme-ified in Spanish.
Per Google Translates, my `anytime` package helped simplify / be an alternate for a tutorial on how to survice Ex#%l date messes.
Really glad it was of help. I use it all the myself too. #rstats
A simpler solution may be to run `update.packages()` daily (or close to it) and to trust both CRAN on the examplary work they do, and stick with packages that won't break you.
There. That "just worked" for me for 25 years with #RStats. These days I even get binaries via r2u.
I have a bad habit of not paying attention to the version of packages I'm using in #Rstats. Then I'm surprised if a package is updated and things break. Want to know how to keep better track of your packages and their versions? Watch the latest Code Club
youtube.com/yc7ZB4F_dc0
New quick `r2u` demo: spawned by the SO question below showing how to install package `raster` plus all its dependencies as binaries in one quick command that run in under 20 seconds. #rstats#rspatialstackoverflow.com/a/72946019…
Hi @RConsortium: Any news on the R Hub Builder status? It is a key service for the #Rstats community and one of you high-profile activies. The complete silence is a little worrysome. Can we get an update?
PSA: @AchimZeileis is a saint. We both noticed yesterday that #Rstats package `BMisc` moved off CRAN for a minor issue ("Suggests not tested) which pulled more packages off with it. But as the man never sleeps here is his extra-clean PR to update `BMisc`:
github.com/bcallaway11/BMisc…
Welcome to town! You could start here: meetup.com/chicago-r-user-gr…
There is a meeting _in two days_ you could attend. Besides the RUG there is a lot more on the ground #Rstats activity, but it is not always aggregated or coordinated.
/cc @troy_phd
This works outside of @Docker too: it is "warp speed" (nice choice of words I may borrow some more) for *any* @Ubuntu-based #RStats deployment. Laptop. Server. Cloud. Whatever.
Timed install of #tidyverse + #shiny (plus helpers) + #Seurat + more earlier today: under one minute.
Notice that I am using the eddelbuettel/r2u based image by @eddelbuettel that brings previously unimaginable warp speed to the docker build process.
github.com/eddelbuettel/r2u#…
Notice that I am using the eddelbuettel/r2u based image by @eddelbuettel that brings previously unimaginable warp speed to the docker build process.
github.com/eddelbuettel/r2u#…
Hi all, unfortunately we temporarily need to stop our Windows and Linux builders, to sort out some billing issues. Hopefully we'll be able to turn them back on soon. We'll tweet here once we have more information.
A PhD-level course in spatial data analysis will be held 14-18 November 2022 in Bergen, Norway. If interested, see: stat.ethz.ch/pipermail/r-sig… for links and details.
prairielearn.org/
There is a active community of users on the PrairieLearn slack instance (which is IIRC open); my course uses an #Rstats autograder container setup described a little in this paper
arxiv.org/abs/2003.06500
Next time anyone talks about arming teachers, remember that today a guy waltzed into downtown Highland Park, shot more than 30 people, and then waltzed out. An entire armed police all around couldn't stop him. Heck, they still can't even find him.