You now know how to declare and build reproducible development
environments using {rix} and Nix. This vignette will
explain how to install specific versions of CRAN packages and how to
install packages from GitHub.
It is important at this stage to understand that you should not call
install.packages() from a running Nix environment. Doing so
will raise an error to avoid issues. If you want to add packages while
analyzing data, you need to add it the default.nix
expression and rebuild the environment. The same goes for installing
packages from GitHub; use the method described in this vignette instead
of using something like remotes::install_github().
We recommend you create a script called create_env.R or
similar, and add the call to rix() there:
library(rix)
rix(r_ver = "4.4.0",
r_pkgs = c("dplyr", "ggplot2"),
system_pkgs = NULL,
git_pkgs = NULL,
ide = "code",
project_path = path_default_nix,
overwrite = TRUE,
print = TRUE)Then, add the packages you need to r_pkgs and run the
script again. Then, build the environment using nix-build
again, and drop into it using nix-shell. Calling
install.packages() is a bad idea for several reasons:
install.packages(), your environment would end up in a
state where the default.nix definition of the environment
and the actual environment don’t match anymore.install.packages() would likely simply not work,
and if it would work, it would cause issues. For example, if you call
install.packages("ggplot2") from one Nix shell, it will not
install {ggplot2} “inside” the Nix shell, but will install
it on your user’s system library of packages. This is because the Nix
shell cannot be changed at run-time, and so, R will instead install the
packages in the user’s library. This version of {ggplot2},
because it is in that system-wide library of packages, will be available
to any other Nix shells. If you call
install.packages("ggplot2") again from another Nix shell,
say 6 months later, this will replace the first version of
{ggplot2} with the latest version.Ideally, you should only manage R versions and R packages using Nix,
and uninstall any system-managed version of R and R packages. But if you
do wish to keep a system-managed version of R and R packages,
rix::rix() also runs rix::rix_init()
automatically which generates an .Rprofile file that avoids
any clashes between your global user library and Nix-managed libraries
of R packages.
It is possible to install an arbitrary version of a package that has been archived on CRAN:
path_default_nix <- tempdir()
rix(
r_ver = "4.2.1",
r_pkgs = c("dplyr@0.8.0", "janitor@1.0.0"),
system_pkgs = NULL,
git_pkgs = NULL,
ide = "other",
project_path = path_default_nix,
overwrite = TRUE
)#> let
#> pkgs = import (fetchTarball "https://github.com/NixOS/nixpkgs/archive/79b3d4bcae8c7007c9fd51c279a8a67acfa73a2a.tar.gz") {};
#>
#> git_archive_pkgs = [
#> (pkgs.rPackages.buildRPackage {
#> name = "dplyr";
#> src = pkgs.fetchzip {
#> url = "https://cran.r-project.org/src/contrib/Archive/dplyr/dplyr_0.8.0.tar.gz";
#> sha256 = "sha256-f30raalLd9KoZKZSxeTN71PG6BczXRIiP6g7EZeH09U=";
#> };
#> propagatedBuildInputs = builtins.attrValues {
#> inherit (pkgs.rPackages)
#> assertthat
#> glue
#> magrittr
#> pkgconfig
#> R6
#> Rcpp
#> rlang
#> tibble
#> tidyselect
#> BH
#> plogr;
#> };
#> })
#>
#> (pkgs.rPackages.buildRPackage {
#> name = "janitor";
#> src = pkgs.fetchzip {
#> url = "https://cran.r-project.org/src/contrib/Archive/janitor/janitor_1.0.0.tar.gz";
#> sha256 = "sha256-3NJomE/CNbOZ+ohuVZJWe2n1RVGUm8x8a0A0qzLmdN4=";
#> };
#> propagatedBuildInputs = builtins.attrValues {
#> inherit (pkgs.rPackages)
#> dplyr
#> tidyr
#> snakecase
#> magrittr
#> purrr
#> rlang;
#> };
#> })
#> ];
#>
#> system_packages = builtins.attrValues {
#> inherit (pkgs)
#> R
#> glibcLocales
#> nix;
#> };
#>
#> in
#>
#> pkgs.mkShell {
#> LOCALE_ARCHIVE = if pkgs.system == "x86_64-linux" then "${pkgs.glibcLocales}/lib/locale/locale-archive" else "";
#> LANG = "en_US.UTF-8";
#> LC_ALL = "en_US.UTF-8";
#> LC_TIME = "en_US.UTF-8";
#> LC_MONETARY = "en_US.UTF-8";
#> LC_PAPER = "en_US.UTF-8";
#> LC_MEASUREMENT = "en_US.UTF-8";
#>
#> buildInputs = [ git_archive_pkgs system_packages ];
#>
#> }The above expression will install R version 4.2.1, and
{dplyr} at version 0.8.0 and {janitor} at
version 1.0.0. This can be useful, especially for packages that have
been archived, but otherwise, this feature should ideally be used
sparingly, for two reasons. First, if you want to reconstruct an
environment as it was around 2019, use the version of R that was current
at that time using the date argument in rix().
This will ensure that every package that gets installed is at a version
compatible with that version of R, which might not be the case if you
need to install a very old version of one particular package. Second,
doing so will install the package from source. For packages that don’t
require compilation, this should be fine, but packages that require
compilation will likely fail to compile successfully. We are working on
handling this better for future versions of {rix}.
It is also possible to install packages from GitHub:
path_default_nix <- tempdir()
rix(
r_ver = "4.2.1",
r_pkgs = c("dplyr", "janitor"),
git_pkgs = list(
list(
package_name = "housing",
repo_url = "https://github.com/rap4all/housing/",
commit = "1c860959310b80e67c41f7bbdc3e84cef00df18e"
),
list(
package_name = "fusen",
repo_url = "https://github.com/ThinkR-open/fusen",
commit = "d617172447d2947efb20ad6a4463742b8a5d79dc"
)
),
ide = "other",
project_path = path_default_nix,
overwrite = TRUE
)This will install two packages from GitHub: the
{housing} package and more specifically the code as it is
in the fusen branch; however the branch name is not
required, as the commit is enough to pin the exact version of the code
needed. The {fusen} package is also installed, as of commit
d617172447d.
It is also possible to install packages from a local
tar.gz file. For this, place the package in the same folder
where the default.nix will be generated, and write
something like this:
rix(
r_ver = "4.3.1",
local_r_pkgs = c("chronicler_0.2.1.tar.gz", "knitr_1.43.tar.gz"),
overwrite = TRUE
)This assumes that both chronicler_0.2.1.tar.gz and
knitr_1.43.tar.gz have been downloaded beforehand. If you
want to include a local package that you are developing, make sure to
first build it using devtools::build() to build the
.tar.gz archive, but if you can, consider uploading the
source code to your package on GitHub and installing it from GitHub
instead.
{rix} also includes an renv2nix() function
that converts an renv.lock file into a valid Nix
expression. Read the vignette vignette("f-renv2nix") to
learn more.
This example shows how to install packages from CRAN, from the CRAN archives and from GitHub:
path_default_nix <- tempdir()
rix(
r_ver = "4.2.1",
r_pkgs = c("dplyr", "janitor", "AER@1.2-8"),
git_pkgs = list(
list(
package_name = "housing",
repo_url = "https://github.com/rap4all/housing/",
commit = "1c860959310b80e67c41f7bbdc3e84cef00df18e"
),
list(
package_name = "fusen",
repo_url = "https://github.com/ThinkR-open/fusen",
commit = "d617172447d2947efb20ad6a4463742b8a5d79dc"
)
),
ide = "other",
project_path = path_default_nix,
overwrite = TRUE
)#> # This file was generated by the {rix} R package v0.7.1 on 2024-07-01
#> # with following call:
#> # >rix(r_ver = "79b3d4bcae8c7007c9fd51c279a8a67acfa73a2a",
#> # > r_pkgs = c("dplyr",
#> # > "janitor",
#> # > "AER@1.2-8"),
#> # > git_pkgs = list(list(package_name = "housing",
#> # > repo_url = "https://github.com/rap4all/housing/",
#> # > commit = "1c860959310b80e67c41f7bbdc3e84cef00df18e"),
#> # > list(package_name = "fusen",
#> # > repo_url = "https://github.com/ThinkR-open/fusen",
#> # > commit = "d617172447d2947efb20ad6a4463742b8a5d79dc")),
#> # > ide = "other",
#> # > project_path = path_default_nix,
#> # > overwrite = TRUE)
#> # It uses nixpkgs' revision 79b3d4bcae8c7007c9fd51c279a8a67acfa73a2a for reproducibility purposes
#> # which will install R version 4.2.1.
#> # Report any issues to https://github.com/ropensci/rix
#> let
#> pkgs = import (fetchTarball "https://github.com/NixOS/nixpkgs/archive/79b3d4bcae8c7007c9fd51c279a8a67acfa73a2a.tar.gz") {};
#>
#> rpkgs = builtins.attrValues {
#> inherit (pkgs.rPackages)
#> dplyr
#> janitor;
#> };
#>
#> git_archive_pkgs = [
#> (pkgs.rPackages.buildRPackage {
#> name = "housing";
#> src = pkgs.fetchgit {
#> url = "https://github.com/rap4all/housing/";
#> rev = "1c860959310b80e67c41f7bbdc3e84cef00df18e";
#> sha256 = "sha256-s4KGtfKQ7hL0sfDhGb4BpBpspfefBN6hf+XlslqyEn4=";
#> };
#> propagatedBuildInputs = builtins.attrValues {
#> inherit (pkgs.rPackages)
#> dplyr
#> ggplot2
#> janitor
#> purrr
#> readxl
#> rlang
#> rvest
#> stringr
#> tidyr;
#> };
#> })
#>
#>
#> (pkgs.rPackages.buildRPackage {
#> name = "fusen";
#> src = pkgs.fetchgit {
#> url = "https://github.com/ThinkR-open/fusen";
#> rev = "d617172447d2947efb20ad6a4463742b8a5d79dc";
#> sha256 = "sha256-TOHA1ymLUSgZMYIA1a2yvuv0799svaDOl3zOhNRxcmw=";
#> };
#> propagatedBuildInputs = builtins.attrValues {
#> inherit (pkgs.rPackages)
#> attachment
#> cli
#> desc
#> devtools
#> glue
#> here
#> magrittr
#> parsermd
#> roxygen2
#> stringi
#> tibble
#> tidyr
#> usethis
#> yaml;
#> };
#> })
#>
#> (pkgs.rPackages.buildRPackage {
#> name = "AER";
#> src = pkgs.fetchzip {
#> url = "https://cran.r-project.org/src/contrib/Archive/AER/AER_1.2-8.tar.gz";
#> sha256 = "sha256-OqxXcnUX/2C6wfD5fuNayc8OU+mstI3tt4eBVGQZ2S0=";
#> };
#> propagatedBuildInputs = builtins.attrValues {
#> inherit (pkgs.rPackages)
#> car
#> lmtest
#> sandwich
#> survival
#> zoo
#> Formula;
#> };
#> })
#> ];
#>
#> system_packages = builtins.attrValues {
#> inherit (pkgs)
#> R
#> glibcLocales
#> nix;
#> };
#>
#> in
#>
#> pkgs.mkShell {
#> LOCALE_ARCHIVE = if pkgs.system == "x86_64-linux" then "${pkgs.glibcLocales}/lib/locale/locale-archive" else "";
#> LANG = "en_US.UTF-8";
#> LC_ALL = "en_US.UTF-8";
#> LC_TIME = "en_US.UTF-8";
#> LC_MONETARY = "en_US.UTF-8";
#> LC_PAPER = "en_US.UTF-8";
#> LC_MEASUREMENT = "en_US.UTF-8";
#>
#> buildInputs = [ git_archive_pkgs rpkgs system_packages ];
#>
#> }The next vignette,
vignette("d2-installing-system-tools-and-texlive-packages-in-a-nix-environment"),
explains how you can install tools such as text editors, git, Quarto,
TexLive packages, and any other tool available through
nixpkgs for your development environments.
Some R packages are quite difficult to install: that is usually not
an issue for most users that use either Windows or macOS as their
operating systems, because when calling install.packages()
a compiled binary gets downloaded from CRAN and installed in a matter of
seconds. On Ubuntu, likely the most popular Linux distribution, binary
packages for R packages are also available via the r2u repository. However,
if you need to install old packages, these instead will need to be
installed from source, as binaries for old packages are not kept. For
most packages, this is not an issue, but some packages require
compilation and this is where issues start.
Nix solves this, because all packages must have their dependencies also declared, so installing old packages should not be an issue. However, it can happen that one particular package that you want to install may not build. This can happen because, even though we spend a lot of time making sure R packages work flawlessly with Nix, there are many R packages (almost 30’000 between CRAN and Bioconductor) and there’s not many of us (R contributors for Nix). Should you have trouble installing a package, feel free to open an issue and we’ll do our best to fix it!
We also made sure that old packages would work by backporting many
fixes, and actually building many old versions of popular packages for
all the dates included in available_dates().
Here is the list of packages that were built and tested (but keep in mind that this list doesn’t show all the dependencies of all the packages that also have to work, and that just because a package isn’t listed, doesn’t mean it’s not going to work!):
DBI
R6
RColorBrewer
RCurl
RSQLite
Rcpp
RcppEigen
arrow
askpass
backports
base64enc
bit
bit64
blob
broom
bslib
cachem
callr
cellranger
cli
clipr
collapse
colorspace
cpp11
crayon
curl
data_table
dbplyr
devtools
digest
dplyr
duckdb
evaluate
fansi
farver
fastmap
fontawesome
forcats
fs
gargle
generics
ggplot2
glue
gtable
haven
highr
hms
htmltools
htmlwidgets
httr
icosa
igraph
isoband
jquerylib
jsonlite
kit
knitr
labeling
languageserver
later
lifecycle
lubridate
magrittr
memoise
mime
modelr
munsell
nloptr
openssl
openxlsx
pillar
pkgconfig
prettyunits
processx
progress
promises
ps
purrr
rJava
ragg
rappdirs
readr
readxl
rematch
rematch2
rlang
rmarkdown
rprojroot
rstan
rstudioapi
rvest
sass
scales
selectr
Seurat
sf
shiny
stars
stringi
stringr
sys
systemfonts
terra
textshaping
tibble
tidyr
tidyselect
tidyverse
timechange
tinytex
tzdb
utf8
vctrs
viridisLite
withr
xfun
xlsx
xml2
yaml
zoo