Data Science. TileDB. Open Source. Quant Research. R. C++. Debian. Linux. Adjunct Clinical Professor, University of Illinois. Lots of coffee. And some running.
Chicago, IL, USA
Joined March 2007
- Tweets 5,361
- Following 350
- Followers 10,149
- Likes 7,646
R^4 #31: Test your package with bleeding edge gcc-11
A recipe, and a Dockerfile and container, to test R packages against the unreleased next gcc/g++ version
dirk.eddelbuettel.com/blog/2…
#rstats #RockerProject @docker
6
1
8
Yep, indeed the fastest JSON parser by some margin.
But credit where credit is due: @knapply_ did all the juicy bits in the #Rstats package building the work by @lemire , @jkeiser2 and others after I had just put together the basic #Rcpp framework.
cloud.r-project.org/package=…
This tweet is unavailable
1
15
RcppTOML 0.1.7 on CRAN: g++-11 support, updates
Better configuration files for humans and machines
dirk.eddelbuettel.com/blog/2…
#rcpp #rstats
3
5
inline 0.3.17 on CRAN: Refactored, and new tests
Inline C, C++ and Fortran programs with ease in your R code
dirk.eddelbuettel.com/blog/2…
Now with much improved ability to store and reload compiled C functions thanks to @JohannesRanke
#rstats
4
1
13
PSA: If you are working on a pull request, and not using the *magic* functionality of @magit_emacs that is `magit-checkout-pull-request` via C-g b y followed by the _pull request issue number_ then ... you are likely doing it wrong.
#rstats
1
2
6
"Caught you" posting and deleting a link to @Infinityssong yesterday--and but I have been binging these folks since. Their covers are pure gold. I'll check out the EP.
youtube.com/playlist?list=PL…
1
Anybody else tired from the marketing drivel in the supposedly community-driven #rstats blog and tweet aggregator?
2
2
Thanks to the Cologne R User Group for having me (alas only virtually with coffee and no Koelsch) to talk @docker, #rstats and #RockerProject. Slides are now up at at dirk.eddelbuettel.com/papers…
/cc @dav_zim
1
3
1
26
Recommended -- that should be good! I will miss it as I will be talking at the same time to the Cologne RUG about @Docker and the #RockerProject -- but I will be sure to check out the recording afterwards.
#rstats
Excited to be speaking about #rstats ALTREP at the @Bioconductor Developer Forum, tomorrow, Thurs Nov 19, 9AM-10AM PST. Talk will be open to the public at bluejeans.com/114067881 Come by if you're interested in learning how the guts of R work.
3
5
The screenshot shows the current version, also released into the drat repo. I cannot control CRAN admittance speed, but I can and do offer (micro)releases via the drat repo. I urge anybody working with ancient compilers to be familiar with these drat repos. Or else help test.
As does version 0.10.1.0.2, released immediately after these issues raised their head. And as stated all it takes to get a micro release from a drat repo is one added parameter : repos="RcppCore.github.io/drat". See the screenshot.
1
1
RcppArmadillo 0.10.1.2.0 on CRAN: A few updates
R bindings to powerful and expressive C++ matrix library
dirk.eddelbuettel.com/blog/2…
#rstats #rcpp
3
1
11
RcppAnnoy 0.0.17: New upstream, new header
(Very Fast) Approximate Nearest Neighbours for R
dirk.eddelbuettel.com/blog/2…
#rstats #rcpp
1
3
15
Another `tidyCpp` shoutout goes to `spdlog` by @gabi_melman: the vignette now has a demo (using `RcppSpdlog`) of the difference between "bad" (assign to `SEXP`) and "good" (instantiate) use of `Protect`. (The `spdlog` init could be common in one file, this was easy ...)
#rstats
1
1
And the new `tidyCpp` release calls for another shoutout for `minidown` by @Atsushi776 and its 'water' vignette style. Crisp yet small documents, allows for side-by-side code snippets from markdown and more, see cloud.r-project.org/web/pack… for an example.
#rstats
2
3
6
tidyCpp 0.0.2 on CRAN: More documentation and features
Tidy C++ wrapping of the C API for R
dirk.eddelbuettel.com/blog/2…
#rstats
1
3
24
RcppSimdJson 0.1.3 on CRAN: New Upstream, New Utilities
Highest-performance JSON Parsing via Modern C++ from R
dirk.eddelbuettel.com/blog/2…
#rcpp #rstats
/cc @knapply_ @lemire @jkeiser2
2
1
18
A refinement to yesterday's revision, correcting to 'which="most", recursive="strong"'.
As @edzerpebesma noted, the recursion makes this very expensive *very quickly*. @MattDowle proposes using only 'which="most"' which seems sensible -- see example.
#rstats
New CRAN Repository Policy rev4334 posted, history at github.com/eddelbuettel/crp/… #rstats
1
3
3
"It depends".
Per email from @MattDowle, for Rcpp it goes from 2105 to 2134 which is marginal. For another package he mentioned: from 30 to 53 which is tolerable.
There may be a hit for 'mostly Suggests' packages like test runners or vignette processors.
#rstats
1