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
CRAN just crossed 13k packages:
R> AP <- available.packages(contrib.url("cran.r-project.org"),filters=list())
R> dim(AP)
[1] 13004 17
R>
#rstats
2
16
2
80
PSA: This looks amazeballs! -- wandbox.org/
"Social compulation" service with
- 45 different C++ compilers up to g++ 9alpha
- a few dozen other languages.
Wow. Sources and more at github.com/melpon/wandbox
1
5
10
You can also work with RcppQuantuccia relying in the header-only Quantuccia subset of QuantLib. Still a heavy compilation (hello, templates!) but without the need for a QL library build. Code may be behind the 1.13.1 calendar code in QL though. PRs welcome.
1
2
Nice one! Particular for the tinyverse.org focus of no dependency.
Maybe a 'message of the day' alike feature would be better? Cache + regenerate if older than, say, 24 hours -- and otherwise cronjob it to avoid slowdown at startup?
Dep-free scriplet now in my .Rprofile to check status of my #rstats CRAN pkgs. Slows R startup a bit, but less disruptive than browseURL version.
gist.github.com/brodieG/e60c…
4
9
Use Open Journals like @jstatsoft / R Journal (@_R_Foundation) / @joss_TheOJ / @ReScienceEds ...
Agreed re okular (KDE pdf reader) which can manage this (even if evince via xdg-open is my goto tool). And once or twice I think I borrowed a Windoze laptop.
1
littler 0.3.4 on CRAN: Extended and improved examples
Everybody's favourite way to rock R at the command-line and in scripts
dirk.eddelbuettel.com/blog/2…
#rstats
1
3
6
digest 0.6.16 on CRAN: robustifcations
Hash digests for arbitrary R objects
dirk.eddelbuettel.com/blog/2…
#rstats
1
4
rfoaas 2.0.0 on CRAN: updated and extended
A RESTful interface for R to the excellent @FOAAS service
dirk.eddelbuettel.com/blog/2…
#rstats
/cc @foaas
1
6
1
11
A nice thing about R/Finance is that you get to meet super-sharp startups which may become sponsors and friend ...
So what do I do now? Rush to ORD to catch a flight to LGA to prevent poor @edouarda14 to convert his _awesome_ @quasardb code to C++98? Oh the puzzles we face ...
To get into Berghain you need to have these kind of deep sad eyes, like you don’t care about life anymore. My trick is to imagine my company got acquired and we have to rewrite everything in C++98 using MFC.
1
5
Our RcppArmadillo package is now used by over 500 CRAN packages!
Congratulations--and thanks!!--to everyone who made it happen: Conrad, Ryan, and everybody on the RcppArmadillo side @romain_francois @BatesDmbates @br1anu and contributors such as @axiomsofxyz
#rcpp #rstats
1
6
1
36
RcppArmadillo 0.9.100.5.0 on CRAN: New major release
Powerful and expressive C++ matrix library
dirk.eddelbuettel.com/blog/2… …
#rcpp #rstats
4
11
Or make it a system-wide setting for all users as eg the @debian package does:
salsa.debian.org/edd/r-base/…
Good piece of advice in there: if you don't like (or need) having a fixed CRAN snapshot with MRO, you can just call:
chooseCRANmirror()
for standard R behavior in this session, or add this to .Rprofile to make it permanent:
options(repos = c(CRAN = “cloud.r-project.org/”))
3
RQuantLib 0.4.5 on CRAN: WIndows binaries are back, small updates
Quantitative Finance functionality for R
dirk.eddelbuettel.com/blog/2…
#rstats #rcpp
With special thanks to @opencpu for the Windows help
/cc @quantlib
5
10
Gem of an answer by @MarkKlik about why his fst package outperforms simpler C/C++ approaches to R object to/from disk serialization (by some margin).
tl;dr: chunk size matters!
stackoverflow.com/a/51716428…
#rstat #rcpp #fst
2
2
37
Reading @yudapearl 's ''Book of Why' noticing scepticism of econometrics so deep-seated that a Nobel-winning concept gets mislabeled: it is vector autoregression hence VAR.
Otherwise mostly nice read so far.
#vacation
1
1
13
I am another R package developer and I endorse this message.
Seriously. Base R rocks. And CRAN is a another jewel. Thank you both.
I try to write low dependency #rstats pkgs, but this is only possible b/c R has so much built-in: stats, string manip, low/high level graphics, etc. And it is stable over time; a bedrock to build on. Me when I think about it: 🤯.
Thank you R-Core!
11
1
100
RcppCNPy 0.2.10 on CRAN: Mostly documentation updates
Read/write NumPy data from R
dirk.eddelbuettel.com/blog/2…
#rcpp #rstats
2
5
This shows a long history of the Tiobe index--and I take the #Rstat growth in here!
Nice article about Python and the recent changes though.
Plus, a Python/R hybrid data analysis here: github.com/TheEconomist/big-… (with a hat tip to @MattDowle)
This tweet is unavailable
3
1
20