Search Results: "Dirk Eddelbuettel"

23 September 2017

Dirk Eddelbuettel: RcppCNPy 0.2.7

A new version of the RcppCNPy package arrived on CRAN yesterday. RcppCNPy provides R with read and write access to NumPy files thanks to the cnpy library by Carl Rogers. This version updates internals for function registration, but otherwise mostly switches the vignette over to the shiny new pinp two-page template and package.

Changes in version 0.2.7 (2017-09-22)
  • Vignette updated to Rmd and use of pinp package
  • File src/init.c added for dynamic registration

CRANberries also provides a diffstat report for the latest release. As always, feedback is welcome and the best place to start a discussion may be the GitHub issue tickets page.

This post by Dirk Eddelbuettel originated on his Thinking inside the box blog. Please report excessive re-aggregation in third-party for-profit settings.

Dirk Eddelbuettel: RcppClassic 0.9.8

A bug-fix release RcppClassic 0.9.8 for the very recent 0.9.7 release which fixes a build issue on macOS introduced in 0.9.7. No other changes. Courtesy of CRANberries, there are changes relative to the previous release. Questions, comments etc should go to the rcpp-devel mailing list off the R-Forge page.

This post by Dirk Eddelbuettel originated on his Thinking inside the box blog. Please report excessive re-aggregation in third-party for-profit settings.

20 September 2017

Dirk Eddelbuettel: pinp 0.0.2: Onwards

A first update 0.0.2 of the pinp package arrived on CRAN just a few days after the initial release. We added a new vignette for the package (see below), extended a few nice features, and smoothed a few corners.
The NEWS entry for this release follows.

Changes in tint version 0.0.2 (2017-09-20)
  • The YAML segment can be used to select font size, one-or-two column mode, one-or-two side mode, linenumbering and watermarks (#21 and #26 addressing #25)
  • If pinp.cls or jss.bst are not present, they are copied in ((#27 addressing #23)
  • Output is now in shaded framed boxen too (#29 addressing #28)
  • Endmatter material is placed in template.tex (#31 addressing #30)
  • Expanded documentation of YAML options in skeleton.Rmd and clarified available one-column option (#32).
  • Section numbering can now be turned on and off (#34)
  • The default bibliography style was changed to jss.bst.
  • A short explanatory vignette was added.

Courtesy of CRANberries, there is a comparison to the previous release. More information is on the tint page. For questions or comments use the issue tracker off the GitHub repo.

This post by Dirk Eddelbuettel originated on his Thinking inside the box blog. Please report excessive re-aggregation in third-party for-profit settings.

17 September 2017

Dirk Eddelbuettel: RcppClassic 0.9.7

A rather boring and otherwise uneventful release 0.9.7 of RcppClassic is now at CRAN. This package provides a maintained version of the otherwise deprecated first Rcpp API; no new projects should use it. Once again no changes in user-facing code. But this makes it the first package to use the very new and shiny pinp package as the backend for its vignette, now converted to Markdown---see here for this new version. We also updated three sources files for tabs versus spaces as the current g++ version complained (correctly !!) about misleading indents. Otherwise a file src/init.c was added for dynamic registration, the Travis CI runner script was updated to using run.sh from our r-travis fork, and we now strip the library after they have been built. Again, no user code changes. And no iterate: nobody should use this package. Rcpp is so much better in so many ways---this one is simply available as we (quite strongly) believe that APIs are contracts, and as such we hold up our end of the deal. Courtesy of CRANberries, there are changes relative to the previous release. Questions, comments etc should go to the rcpp-devel mailing list off the R-Forge page.

This post by Dirk Eddelbuettel originated on his Thinking inside the box blog. Please report excessive re-aggregation in third-party for-profit settings.

16 September 2017

Dirk Eddelbuettel: pinp 0.0.1: pinp is not PNAS

A brandnew and very exciting (to us, at least) package called pinp just arrived on CRAN, following a somewhat unnecessarily long passage out of incoming. It is based on the PNAS LaTeX Style offered by the Proceeding of the National Academy of Sciences of the United States of America, or PNAS for short. And there is already a Markdown version in the wonderful rticles packages. But James Balamuta and I thought we could do one better when we were looking to typeset our recent PeerJ Prepint as an attractive looking vignette for use within the Rcpp package. And so we did by changing a few things (font, color, use of natbib and Chicago.bst for references, removal of a bunch of extra PNAS-specific formalities from the frontpage) and customized a number of other things for easier use by vignettes directly from the YAML header (draft mode watermark, doi or url for packages, easier author naming in footer, bibtex file and more). We are quite pleased with the result which seems ready for the next Rcpp release---see e.g., these two teasers:
and
and the pinp package page or the GitHub repo have the full (four double-)pages of what turned a more dull looking 27 page manuscript into eight crisp two-column pages. We have few more things planned (i.e., switching to single column mode, turning on linenumbers at least in one-column mode). For questions or comments use the issue tracker off the GitHub repo.

This post by Dirk Eddelbuettel originated on his Thinking inside the box blog. Please report excessive re-aggregation in third-party for-profit settings.

Dirk Eddelbuettel: drat 0.1.3

A new version of drat arrived earlier today on CRAN as another no-human-can-delay-this automatic upgrade directly from the CRAN prechecks. It is mostly a maintenance release ensuring PACKAGES.rds is also updated, plus some minor other edits. drat stands for drat R Archive Template, and helps with easy-to-create and easy-to-use repositories for R packages. Since its inception in early 2015 it has found reasonably widespread adoption among R users because repositories with marked releases is the better way to distribute code. The NEWS file summarises the release as follows:

Changes in drat version 0.1.3 (2017-09-16)
  • Ensure PACKAGES.rds, if present, is also inserted in repo
  • Updated 'README.md' removing stale example URLs (#63)
  • Use https to fetch Travis CI script from r-travis

Courtesy of CRANberries, there is a comparison to the previous release. More detailed information is on the drat page.

This post by Dirk Eddelbuettel originated on his Thinking inside the box blog. Please report excessive re-aggregation in third-party for-profit settings.

14 September 2017

Dirk Eddelbuettel: RcppMsgPack 0.2.0

A new and much enhanced version of RcppMsgPack arrived on CRAN a couple of days ago. It came together following this email to the r-package-devel list which made it apparent that Travers Ching had been working on MessagePack converters for R which required the very headers I had for use from, inter alia, the RcppRedis package. So we joined our packages. I updated the headers in RcppMsgPack to the current upstream version 2.1.5 of MessagePack, and Travers added his helper functions allow direct packing / unpacking of MessagePack objects at the R level, as well as tests and a draft vignette. Very exciting, and great to have a coauthor! So now RcppMspPack provides R with both MessagePack header files for use via C++ (or C, if you must) packages such as RcppRedis --- and direct conversion routines at the R prompt. MessagePack itself is an efficient binary serialization format. It lets you exchange data among multiple languages like JSON. But it is faster and smaller. Small integers are encoded into a single byte, and typical short strings require only one extra byte in addition to the strings themselves.

Changes in version 0.2.0 (2017-09-07)
  • Added support for building on Windows
  • Upgraded to MsgPack 2.1.5 (#3)
  • New R functions to manipulate MsgPack objects: msgpack_format, msgpack_map, msgpack_pack, msgpack_simplify, mgspack_unpack (#4)
  • New R functions also available as msgpackFormat, msgpackMap, msgpackPack, msgpackSimplify, mgspackUnpack (#4)
  • New vignette (#4)
  • New tests (#4)

Courtesy of CRANberries, there is also a diffstat report for this release. More information is on the RcppRedis page. More information may be on the RcppMsgPack page. Issues and bugreports should go to the GitHub issue tracker.

This post by Dirk Eddelbuettel originated on his Thinking inside the box blog. Please report excessive re-aggregation in third-party for-profit settings.

Dirk Eddelbuettel: RcppRedis 0.1.8

A new minor release of RcppRedis arrived on CRAN last week, following the release 0.2.0 of RcppMsgPack which brought the MsgPack headers forward to release 2.1.5. This required a minor and rather trivial change in the code. When the optional RcppMsgPack package is used, we now require this version 0.2.0 or later. We made a few internal updates to the package as well.

Changes in version 0.1.8 (2017-09-08)
  • A new file init.c was added with calls to R_registerRoutines() and R_useDynamicSymbols()
  • Symbol registration is enabled in useDynLib
  • Travis CI was updated to using run.sh
  • The (optional MessagePack) code was updated for MsgPack 2.*

Courtesy of CRANberries, there is also a diffstat report for this release. More information is on the RcppRedis page.

This post by Dirk Eddelbuettel originated on his Thinking inside the box blog. Please report excessive re-aggregation in third-party for-profit settings.

1 September 2017

Dirk Eddelbuettel: RcppAnnoy 0.0.9

An new version 0.0.9 of RcppAnnoy, our Rcpp-based R integration of the nifty Annoy library by Erik, is now on CRAN. Annoy is a small and lightweight C++ template header library for very fast approximate nearest neighbours. This release corrects an issue for Windows users discovered by GitHub user 'khoran' who later also suggested the fix of binary mode. It upgrades to Annoy release 1.9.1 and brings its new Manhattan distance to RcppAnnoy. A number of unit tests were added as well, and we updated some packaging internals such as symbol registration. And I presume I had a good streak emailing with Uwe's robots as the package made it onto CRAN rather smoothly within ten minutes of submission:
RcppAnnou to CRAN
Changes in this version are summarized here:

Changes in version 0.0.9 (2017-08-31)
  • Synchronized with Annoy upstream version 1.9.1
  • Minor updates in calls and tests as required by annoy 1.9.1
  • New Manhattan distance modules along with unit test code
  • Additional unit tests from upstream test code carried over
  • Binary mode is used for save (as suggested by @khoran in #21)
  • A new file init.c was added with calls to R_registerRoutines() and R_useDynamicSymbols()
  • Symbol registration is enabled in useDynLib

Courtesy of CRANberries, there is also a diffstat report for this release.

This post by Dirk Eddelbuettel originated on his Thinking inside the box blog. Please report excessive re-aggregation in third-party for-profit settings.

30 August 2017

Dirk Eddelbuettel: RcppArmadillo 0.7.960.1.2

armadillo image A second fix-up release is needed following on the recent bi-monthly RcppArmadillo release as well as the initial follow-up as it turns out that OS X / macOS is so darn special that it needs an entire separate treatment for OpenMP. Namely to turn it off entirely... Armadillo is a powerful and expressive C++ template library for linear algebra aiming towards a good balance between speed and ease of use with a syntax deliberately close to a Matlab. RcppArmadillo integrates this library with the R environment and language--and is widely used by (currently) 384 other packages on CRAN---an increase of 54 since the CRAN release in June!

Changes in RcppArmadillo version 0.7.960.1.2 (2017-08-29)
  • On macOS, OpenMP support is now turned off (#170).
  • The package is now compiling under the C++11 standard (#170).
  • The vignette dependency is correctly set (James and Dirk in #168 and #169)

Courtesy of CRANberries, there is a diffstat report. More detailed information is on the RcppArmadillo page. Questions, comments etc should go to the rcpp-devel mailing list off the R-Forge page.

This post by Dirk Eddelbuettel originated on his Thinking inside the box blog. Please report excessive re-aggregation in third-party for-profit settings.

29 August 2017

Dirk Eddelbuettel: RcppSMC 0.2.0

A new version 0.2.0 of the RcppSMC package arrived on CRAN earlier today (as a very quick pretest-publish within minutes of submission). RcppSMC provides Rcpp-based bindings to R for the Sequential Monte Carlo Template Classes (SMCTC) by Adam Johansen described in his JSS article. This release 0.2.0 is chiefly the work of Leah South, a Ph.D. student at Queensland University of Technology, who was during the last few months a Google Summer of Code student mentored by Adam and myself. It was pleasure to work with Leah on this, and see her progress. Our congratulations to Leah for a job well done!

Changes in RcppSMC version 0.2.0 (2017-08-28)
  • Also use .registration=TRUE in useDynLib in NAMESPACE
  • Multiple Sequential Monte Carlo extensions (Leah South as part of Google Summer of Code 2017)
    • Switching to population level objects (#2 and #3).
    • Using Rcpp attributes (#2).
    • Using automatic RNGscope (#4 and #5).
    • Adding multiple normalising constant estimators (#7).
    • Static Bayesian model example: linear regression (#10 addressing #9).
    • Adding a PMMH example (#13 addressing #11).
    • Framework for additional algorithm parameters and adaptation (#19 addressing #16; also #24 addressing #23).
    • Common adaptation methods for static Bayesian models (#20 addressing #17).
    • Supporting MCMC repeated runs (#21).
    • Adding adaptation to linear regression example (#22 addressing #18).

Courtesy of CRANberries, there is a diffstat report for this release. More information is on the RcppSMC page. Issues and bugreports should go to the GitHub issue tracker.

This post by Dirk Eddelbuettel originated on his Thinking inside the box blog. Please report excessive re-aggregation in third-party for-profit settings.

25 August 2017

Dirk Eddelbuettel: BH 1.65.0-1

The BH package on CRAN was updated today to version 1.65.0. BH provides a sizeable portion of the Boost C++ libraries as a set of template headers for use by R, possibly with Rcpp as well as other packages. This release upgrades the version of Boost to the rather new upstream version Boost 1.65.0 released earlier this week, and adds two new libraries: align and sort. I had started the upgrade process a few days ago under release 1.64.0. Rigorous checking of reverse dependencies showed that mvnfast needed a small change (which was trivial: just seeding the RNG prior to running tests), which Matteo did in no time with a fresh CRAN upload. rstan is needing a bit more work but should be ready real soon now and we are awaiting a new version. And once I switched to the just release Boost 1.65.0 it became apparent that Cyclops no longer needs its embedded copy of Boost iterator---and Marc already made that change with yet another fresh CRAN upload. It is a true pleasure to work in such a responsive and collaborative community.

Changes in version 1.65.0-1 (2017-08-24)
  • Upgraded to Boost 1.64 and then 1.65 installed directly from upstream source with several minor tweaks (as before)
  • Fourth tweak corrects a misplaced curly brace (see the Boost ublas GitHub repo and its issue #40)
  • Added Boost align (as requested in #32)
  • Added Boost sort (as requested in #35)
  • Added Boost multiprecision by fixing a script typo (as requested in #42)
  • Updated Travis CI support via newer run.sh

Via CRANberries, there is a diffstat report relative to the previous release. Comments and suggestions are welcome via the mailing list or the issue tracker at the GitHub repo.

This post by Dirk Eddelbuettel originated on his Thinking inside the box blog. Please report excessive re-aggregation in third-party for-profit settings.

23 August 2017

Dirk Eddelbuettel: Rcpp now used by 10 percent of CRAN packages

10 percent of CRAN packages Over the last few days, Rcpp passed another noteworthy hurdle. It is now used by over 10 percent of packages on CRAN (as measured by Depends, Imports and LinkingTo, but excluding Suggests). As of this morning 1130 packages use Rcpp out of a total of 11275 packages. The graph on the left shows the growth of both outright usage numbers (in darker blue, left axis) and relative usage (in lighter blue, right axis). Older posts on this blog took note when Rcpp passed round hundreds of packages, most recently in April for 1000 packages. The growth rates for both Rcpp, and of course CRAN, are still staggering. A big thank you to everybody who makes this happen, from R Core and CRAN to all package developers, contributors, and of course all users driving this. We have built ourselves a rather impressive ecosystem. So with that a heartfelt Thank You! to all users and contributors of R, CRAN, and of course Rcpp, for help, suggestions, bug reports, documentation, encouragement, and, of course, code.

This post by Dirk Eddelbuettel originated on his Thinking inside the box blog. Please report excessive re-aggregation in third-party for-profit settings.

20 August 2017

Dirk Eddelbuettel: RcppArmadillo 0.7.960.1.1

armadillo image On the heels of the very recent bi-monthly RcppArmadillo release comes a quick bug-fix release 0.7.960.1.1 which just got onto CRAN (and I will ship a build to Debian in a moment). There were three distinct issues I addressed in three quick pull requests: Armadillo is a powerful and expressive C++ template library for linear algebra aiming towards a good balance between speed and ease of use with a syntax deliberately close to a Matlab. RcppArmadillo integrates this library with the R environment and language--and is widely used by (currently) 382 other packages on CRAN---an increase of 52 since the CRAN release in June! Changes in this release relative to the previous CRAN release are as follows:

Changes in RcppArmadillo version 0.7.960.1.1 (2017-08-20)
  • Added improved check for inherited S4 matrix classes (#162 fixing #161)
  • Changed fastLm C++ function to fastLm_impl to not clash with R method (#164 fixing #163)
  • Added OpenMP check for configure (#166 fixing #165)

Courtesy of CRANberries, there is a diffstat report. More detailed information is on the RcppArmadillo page. Questions, comments etc should go to the rcpp-devel mailing list off the R-Forge page.

This post by Dirk Eddelbuettel originated on his Thinking inside the box blog. Please report excessive re-aggregation in third-party for-profit settings.

Dirk Eddelbuettel: #10: Compacting your Shared Libraries, After The Build

Welcome to the tenth post in the rarely ranting R recommendations series, or R4 for short. A few days ago we showed how to tell the linker to strip shared libraries. As discussed in the post, there are two options. One can either set up ~/.R/Makevars by passing the strip-debug option to the linker. Alternatively, one can adjust src/Makevars in the package itself with a bit a Makefile magic. Of course, there is a third way: just run strip --strip-debug over all the shared libraries after the build. As the path is standardized, and the shell does proper globbing, we can just do
$ strip --strip-debug /usr/local/lib/R/site-library/*/libs/*.so
using a double-wildcard to get all packages (in that R package directory) and all their shared libraries. Users on macOS probably want .dylib on the end, users on Windows want another computer as usual (just kidding: use .dll). Either may have to adjust the path which is left as an exercise to the reader. The impact can be Yuge as illustrated in the following dotplot:
This illustration is in response to a mailing list post. Last week, someone claimed on r-help that tidyverse would not install on Ubuntu 17.04. And this is of course patently false as many of us build and test on Ubuntu and related Linux systems, Travis runs on it, CRAN tests them etc pp. That poor user had somehow messed up their default gcc version. Anyway: I fired up a Docker container, installed r-base-core plus three required -dev packages (for xml2, openssl, and curl) and ran a single install.packages("tidyverse"). In a nutshell, following the launch of Docker for an Ubuntu 17.04 container, it was just
$ apt-get update
$ apt-get install r-base libcurl4-openssl-dev libssl-dev libxml2-dev
$ apt-get install mg          # a tiny editor
$ mg /etc/R/Rprofile.site     # to add a default CRAN repo
$ R -e 'install.packages("tidyverse")'
which not only worked (as expected) but also installed a whopping fifty-one packages (!!) of which twenty-six contain a shared library. A useful little trick is to run du with proper options to total, summarize, and use human units which reveals that these libraries occupy seventy-eight megabytes:
root@de443801b3fc:/# du -csh /usr/local/lib/R/site-library/*/libs/*so
4.3M    /usr/local/lib/R/site-library/Rcpp/libs/Rcpp.so
2.3M    /usr/local/lib/R/site-library/bindrcpp/libs/bindrcpp.so
144K    /usr/local/lib/R/site-library/colorspace/libs/colorspace.so
204K    /usr/local/lib/R/site-library/curl/libs/curl.so
328K    /usr/local/lib/R/site-library/digest/libs/digest.so
33M     /usr/local/lib/R/site-library/dplyr/libs/dplyr.so
36K     /usr/local/lib/R/site-library/glue/libs/glue.so
3.2M    /usr/local/lib/R/site-library/haven/libs/haven.so
272K    /usr/local/lib/R/site-library/jsonlite/libs/jsonlite.so
52K     /usr/local/lib/R/site-library/lazyeval/libs/lazyeval.so
64K     /usr/local/lib/R/site-library/lubridate/libs/lubridate.so
16K     /usr/local/lib/R/site-library/mime/libs/mime.so
124K    /usr/local/lib/R/site-library/mnormt/libs/mnormt.so
372K    /usr/local/lib/R/site-library/openssl/libs/openssl.so
772K    /usr/local/lib/R/site-library/plyr/libs/plyr.so
92K     /usr/local/lib/R/site-library/purrr/libs/purrr.so
13M     /usr/local/lib/R/site-library/readr/libs/readr.so
4.7M    /usr/local/lib/R/site-library/readxl/libs/readxl.so
1.2M    /usr/local/lib/R/site-library/reshape2/libs/reshape2.so
160K    /usr/local/lib/R/site-library/rlang/libs/rlang.so
928K    /usr/local/lib/R/site-library/scales/libs/scales.so
4.9M    /usr/local/lib/R/site-library/stringi/libs/stringi.so
1.3M    /usr/local/lib/R/site-library/tibble/libs/tibble.so
2.0M    /usr/local/lib/R/site-library/tidyr/libs/tidyr.so
1.2M    /usr/local/lib/R/site-library/tidyselect/libs/tidyselect.so
4.7M    /usr/local/lib/R/site-library/xml2/libs/xml2.so
78M     total
root@de443801b3fc:/# 
Looks like dplyr wins this one at thirty-three megabytes just for its shared library. But with a single stroke of strip we can reduce all this down a lot:
root@de443801b3fc:/# strip --strip-debug /usr/local/lib/R/site-library/*/libs/*so
root@de443801b3fc:/# du -csh /usr/local/lib/R/site-library/*/libs/*so
440K    /usr/local/lib/R/site-library/Rcpp/libs/Rcpp.so
220K    /usr/local/lib/R/site-library/bindrcpp/libs/bindrcpp.so
52K     /usr/local/lib/R/site-library/colorspace/libs/colorspace.so
56K     /usr/local/lib/R/site-library/curl/libs/curl.so
120K    /usr/local/lib/R/site-library/digest/libs/digest.so
2.5M    /usr/local/lib/R/site-library/dplyr/libs/dplyr.so
16K     /usr/local/lib/R/site-library/glue/libs/glue.so
404K    /usr/local/lib/R/site-library/haven/libs/haven.so
76K     /usr/local/lib/R/site-library/jsonlite/libs/jsonlite.so
20K     /usr/local/lib/R/site-library/lazyeval/libs/lazyeval.so
24K     /usr/local/lib/R/site-library/lubridate/libs/lubridate.so
8.0K    /usr/local/lib/R/site-library/mime/libs/mime.so
52K     /usr/local/lib/R/site-library/mnormt/libs/mnormt.so
84K     /usr/local/lib/R/site-library/openssl/libs/openssl.so
76K     /usr/local/lib/R/site-library/plyr/libs/plyr.so
32K     /usr/local/lib/R/site-library/purrr/libs/purrr.so
648K    /usr/local/lib/R/site-library/readr/libs/readr.so
400K    /usr/local/lib/R/site-library/readxl/libs/readxl.so
128K    /usr/local/lib/R/site-library/reshape2/libs/reshape2.so
56K     /usr/local/lib/R/site-library/rlang/libs/rlang.so
100K    /usr/local/lib/R/site-library/scales/libs/scales.so
496K    /usr/local/lib/R/site-library/stringi/libs/stringi.so
124K    /usr/local/lib/R/site-library/tibble/libs/tibble.so
164K    /usr/local/lib/R/site-library/tidyr/libs/tidyr.so
104K    /usr/local/lib/R/site-library/tidyselect/libs/tidyselect.so
344K    /usr/local/lib/R/site-library/xml2/libs/xml2.so
6.6M    total
root@de443801b3fc:/#
Down to six point six megabytes. Not bad for one command. The chart visualizes the respective reductions. Clearly, C++ packages (and their template use) lead to more debugging symbols than plain old C code. But once stripped, the size differences are not that large. And just to be plain, what we showed previously in post #9 does the same, only already at installation stage. The effects are not cumulative.

This post by Dirk Eddelbuettel originated on his Thinking inside the box blog. Please report excessive re-aggregation in third-party for-profit settings.

18 August 2017

Dirk Eddelbuettel: RcppArmadillo 0.7.960.1.0

armadillo image The bi-monthly RcppArmadillo release is out with a new version 0.7.960.1.0 which is now on CRAN, and will get to Debian in due course. And it is a big one. Lots of nice upstream changes from Armadillo, and lots of work on our end as the Google Summer of Code project by Binxiang Ni, plus a few smaller enhancements -- see below for details. Armadillo is a powerful and expressive C++ template library for linear algebra aiming towards a good balance between speed and ease of use with a syntax deliberately close to a Matlab. RcppArmadillo integrates this library with the R environment and language--and is widely used by (currently) 379 other packages on CRAN---an increase of 49 since the last CRAN release in June! Changes in this release relative to the previous CRAN release are as follows:

Changes in RcppArmadillo version 0.7.960.1.0 (2017-08-11)
  • Upgraded to Armadillo release 7.960.1 (Northern Banana Republic Deluxe)
    • faster randn() when using OpenMP (NB: usually omitted when used fromR)
    • faster gmm_diag class, for Gaussian mixture models with diagonal covariance matrices
    • added .sum_log_p() to the gmm_diag class
    • added gmm_full class, for Gaussian mixture models with full covariance matrices
    • expanded .each_slice() to optionally use OpenMP for multi-threaded execution
  • Upgraded to Armadillo release 7.950.0 (Northern Banana Republic)
    • expanded accu() and sum() to use OpenMP for processing expressions with computationally expensive element-wise functions
    • expanded trimatu() and trimatl() to allow specification of the diagonal which delineates the boundary of the triangular part
  • Enhanced support for sparse matrices (Binxiang Ni as part of Google Summer of Code 2017)
    • Add support for dtCMatrix and dsCMatrix (#135)
    • Add conversion and unit tests for dgT, dtT and dsTMatrix (#136)
    • Add conversion and unit tests for dgR, dtR and dsRMatrix (#139)
    • Add conversion and unit tests for pMatrix and ddiMatrix (#140)
    • Rewrite conversion for dgT, dtT and dsTMatrix, and add file-based tests (#142)
    • Add conversion and unit tests for indMatrix (#144)
    • Rewrite conversion for ddiMatrix (#145)
    • Add a warning message for matrices that cannot be converted (#147)
    • Add new vignette for sparse matrix support (#152; Dirk in #153)
    • Add support for sparse matrix conversion from Python SciPy (#158 addressing #141)
  • Optional return of row or column vectors in collapsed form if appropriate #define is set (Serguei Sokol in #151 and #154)
  • Correct speye() for non-symmetric cases (Qiang Kou in #150 closing #149).
  • Ensure tests using Scientific Python and reticulate are properly conditioned on the packages being present.
  • Added .aspell/ directory with small local directory now supported by R-devel.

Courtesy of CRANberries, there is a diffstat report. More detailed information is on the RcppArmadillo page. Questions, comments etc should go to the rcpp-devel mailing list off the R-Forge page.

This post by Dirk Eddelbuettel originated on his Thinking inside the box blog. Please report excessive re-aggregation in third-party for-profit settings.

15 August 2017

Dirk Eddelbuettel: #9: Compacting your Shared Libraries

Welcome to the nineth post in the recognisably rancid R randomness series, or R4 for short. Following on the heels of last week's post, we aim to look into the shared libraries created by R. We love the R build process. It is robust, cross-platform, reliable and rather predicatable. It. Just. Works. One minor issue, though, which has come up once or twice in the past is the (in)ability to fully control all compilation options. R will always recall CFLAGS, CXXFLAGS, ... etc as used when it was compiled. Which often entails the -g flag for debugging which can seriously inflate the size of the generated object code. And once stored in $ RHOME /etc/Makeconf we cannot on the fly override these values. But there is always a way. Sometimes even two. The first is local and can be used via the (personal) ~/.R/Makevars file (about which I will have to say more in another post). But something I have been using quite a bite lately uses the flags for the shared library linker. Given that we can have different code flavours and compilation choices---between C, Fortran and the different C++ standards---one can end up with a few lines. I currently use this which uses -Wl, to pass an the -S (or --strip-debug) option to the linker (and also reiterates the desire for a shared library, presumably superfluous):
SHLIB_CXXLDFLAGS = -Wl,-S -shared
SHLIB_CXX11LDFLAGS = -Wl,-S -shared
SHLIB_CXX14LDFLAGS = -Wl,-S -shared
SHLIB_FCLDFLAGS = -Wl,-S -shared
SHLIB_LDFLAGS = -Wl,-S -shared
Let's consider an example: my most recently uploaded package RProtoBuf. Built under a standard 64-bit Linux setup (Ubuntu 17.04, g++ 6.3) and not using the above, we end up with library containing 12 megabytes (!!) of object code:
edd@brad:~/git/rprotobuf(feature/fewer_warnings)$ ls -lh src/RProtoBuf.so
-rwxr-xr-x 1 edd edd 12M Aug 14 20:22 src/RProtoBuf.so
edd@brad:~/git/rprotobuf(feature/fewer_warnings)$ 
However, if we use the flags shown above in .R/Makevars, we end up with much less:
edd@brad:~/git/rprotobuf(feature/fewer_warnings)$ ls -lh src/RProtoBuf.so 
-rwxr-xr-x 1 edd edd 626K Aug 14 20:29 src/RProtoBuf.so
edd@brad:~/git/rprotobuf(feature/fewer_warnings)$ 
So we reduced the size from 12mb to 0.6mb, an 18-fold decrease. And the file tool still shows the file as 'not stripped' as it still contains the symbols. Only debugging information was removed. What reduction in size can one expect, generally speaking? I have seen substantial reductions for C++ code, particularly when using tenmplated code. More old-fashioned C code will be less affected. It seems a little difficult to tell---but this method is my new build default as I continually find rather substantial reductions in size (as I tend to work mostly with C++-based packages). The second option only occured to me this evening, and complements the first which is after all only applicable locally via the ~/.R/Makevars file. What if we wanted it affect each installation of a package? The following addition to its src/Makevars should do:
strippedLib: $(SHLIB)
        if test -e "/usr/bin/strip"; then /usr/bin/strip --strip-debug $(SHLIB); fi
.phony: strippedLib
We declare a new Makefile target strippedLib. But making it dependent on $(SHLIB), we ensure the standard target of this Makefile is built. And by making the target .phony we ensure it will always be executed. And it simply tests for the strip tool, and invokes it on the library after it has been built. Needless to say we get the same reduction is size. And this scheme may even pass muster with CRAN, but I have not yet tried. Lastly, and acknowledgement. Everything in this post has benefited from discussion with my former colleague Dan Dillon who went as far as setting up tooling in his r-stripper repository. What we have here may be simpler, but it would not have happened with what Dan had put together earlier.

This post by Dirk Eddelbuettel originated on his Thinking inside the box blog. Please report excessive re-aggregation in third-party for-profit settings.

13 August 2017

Dirk Eddelbuettel: RProtoBuf 0.4.10

RProtoBuf provides R bindings for the Google Protocol Buffers ("ProtoBuf") data encoding and serialization library used and released by Google, and deployed fairly widely in numerous projects as a language and operating-system agnostic protocol. A new releases RProtoBuf 0.4.10 just appeared on CRAN. It is a maintenance releases replacing one leftover errorenous use of package= in .Call with the correct PACKAGE= (as requsted by CRAN). It also integrates a small robustification in the deserializer when encountering invalide objects; this was both reported and fixed by Jeffrey Shen.

Changes in RProtoBuf version 0.4.10 (2017-08-13)
  • More careful operation in deserializer checking for a valid class attribute (Jeffrey Shen in #29 fixing #28)
  • At the request of CRAN, correct one .Call() argument to PACKAGE=; update routine registration accordingly

CRANberries also provides a diff to the previous release. The RProtoBuf page has an older package vignette, a 'quick' overview vignette, a unit test summary vignette, and the pre-print for the JSS paper. Questions, comments etc should go to the GitHub issue tracker off the GitHub repo.

This post by Dirk Eddelbuettel originated on his Thinking inside the box blog. Please report excessive re-aggregation in third-party for-profit settings.

11 August 2017

Dirk Eddelbuettel: #8: Customizing Spell Checks for R CMD check

Welcome to the eight post in the ramblingly random R rants series, or R4 for short. We took a short break over the last few weeks due to some conferencing followed by some vacationing and general chill. But we're back now, and this post gets us back to initial spirit of (hopefully) quick and useful posts. Perusing yesterday's batch of CRANberries posts, I noticed a peculiar new directory shown the in the diffstat output we use to compare two subsequent source tarballs. It was entitled .aspell/, in the top-level directory, and in two new packages by R Core member Kurt Hornik himself. The context is, of course, the not infrequently-expressed desire to customize the spell checking done on CRAN incoming packages, see e.g. this r-package-devel thread. And now we can as I verified with (the upcoming next release of) RcppArmadillo, along with a recent-enough (i.e. last few days) version of r-devel. Just copying what Kurt did, i.e. adding a file .aspell/defaults.R, and in it pointing to rds file (named as the package) containing a character vector with words added to the spell checker's universe is all it takes. For my package, see here for the peculiars. Or see here:
edd@bud:~/git/rcpparmadillo/.aspell(master)$ cat defaults.R 
Rd_files <- vignettes <- R_files <- description <-
    list(encoding = "UTF-8",
         language = "en",
         dictionaries = c("en_stats", "RcppArmadillo"))
edd@bud:~/git/rcpparmadillo/.aspell(master)$ r -p -e 'readRDS("RcppArmadillo.rds")'
[1] "MPL"            "Sanderson"      "Templated"
[4] "decompositions" "onwards"        "templated"
edd@bud:~/git/rcpparmadillo/.aspell(master)$     
And now R(-devel) CMD check --as-cran ... is silent about spelling. Yay! But take this with a grain of salt as this does not yet seem to be "announced" as e.g. yesterday's change in the CRAN Policy did not mention it. So things may well change -- but hey, it worked for me. And this all is about aspell, here is something topical about a spell to close the post:

This post by Dirk Eddelbuettel originated on his Thinking inside the box blog. Please report excessive re-aggregation in third-party for-profit settings.

4 August 2017

Dirk Eddelbuettel: R for System Adminstration

Just getting back from the most fun meetup I have been to in quite some time: episode 23 (by their count) of Open Source Open Mic hosted by Matt Godbolt and Joe Walnes here in Chicago. Nothing but a sequence of lightning talks. Plus beer and pizza. Sounds awesome? It was! We had fantastic talks across at least half a dozen languages, covering both new-ish (Pony) and interesting ones such (Rust, Go, ...) plus of course some Javascript and some Python, no Java (yay!) and a few batshit crazy things like a self-hosting database in its own (shell) code, a terminal gif viewer (!!), and more. And it gave me an opportunity to quickly (one evening and morning commute) jam out a presentation about what is in the title: R for system administration. And I am only half-joking. I had used R a couple of years ago when I needed to select, subset, modify, ... a large number of image files given some timestamp and filename patterns. And given how well R works in a vectorised manner with both regular expressions and timestamps, as well as on top of essentially all standard POSIX-style operating system / file-system functions, I picked up that thread again on the problem of ... cleaning up the file storage underlying CRANberries which by now has well over fifty-seven thousand (!!) tarballs of CRAN packages based on now ten years of CRANberries. So I showed how to prune this in essentially half a dozen lines of R (and data.table code), plus some motivation---all just right for a lightning talk. Seemingly the talk went well enough as quite a few folks gave a thumbs up and compliments over beers afterwards. But see for yourself as the slides are now uploaded to my standard talks page. My thanks to Matt and Joe for organizing the meetup. I think I will be back.

Next.

Previous.