Updating to R 3.6.0 on MacOS
Mixed emotions
Every time there is a new major update from The R Foundation (like the recent 3.6.0 release in April). I’m always happy to see the continuing progress and the combination of new features and bug fixes, but I also dread the upgrade because it means I have to address the issue of what to do about the burgeoning number of packages (libraries) I have installed.
Up until now I confess I simply have sort of “winged it”, done the upgrade and either manually thought about what packages I “really” needed or just grabbed a few essentials and then let my needs dictate whatever else I reloaded. This time I decided to get serious about the process and pay attention to not only what I was doing but documenting it and keeping a record via some amount of coding (and this post).
I’m aware that there are full-fledged package
managers like
packrat and checkpoint and even a package designed to manage the upgrade for
you on windows, but I’m a Mac user and wanted to do things my own way and I
don’t need that level of sophistication.
So I set out to do the following:
- Capture a list of everything I had installed under
R 3.5.3and, very importantly, as much as I could about where I got the package e.g.CRANorGitHubor ??? - Keep a copy for my own edification and potential future use.
- Do a clean
R 3.6.0install and not copy any library directories manually. - Take a look at the list I produced in #1 above but mainly to just download and install the exact same packages if I can find them.
- Make the process mainly scripted and automatic and available again for the future.
Helpful background
As I was searching the web I found a few helpful posts that saved me time in
building my own solution. The primary was this
post
on Stack Overflow. I wanted to extend the function listed there to do a little
more of my work for me. Instead of just being able to generate a listing of what
I had installed from GitHub I wanted to be able to determine most of the places
I get packages from, which are CRAN, GitHub and R-Forge.
So let’s load tidyverse to have access to all it’s various functions and
features and then build a dataframe called allmypackages with the basic
information about the packages I currently have installed in R 3.5.3.
Note - I’m writing this after already upgrading so there will be a few inconsistencies in the output
- This could just as easily be a
tibblebut I choseas.data.frame - I am deliberately removing base packages from the dataframe by
filter - I am eliminating columns I really don’t care about with
select
require(tidyverse)
allmypackages <- as.data.frame(installed.packages())
allmypackages <- allmypackages %>%
filter(Priority != "base" | is.na(Priority)) %>%
select(-c(Enhances:MD5sum, LinkingTo:Suggests)) %>%
droplevels()
str(allmypackages)## 'data.frame': 718 obs. of 8 variables:
## $ Package : chr "abind" "acepack" "AER" "afe" ...
## $ LibPath : chr "/Library/Frameworks/R.framework/Versions/4.0/Resources/library" "/Library/Frameworks/R.framework/Versions/4.0/Resources/library" "/Library/Frameworks/R.framework/Versions/4.0/Resources/library" "/Library/Frameworks/R.framework/Versions/4.0/Resources/library" ...
## $ Version : chr "1.4-5" "1.4.1" "1.2-9" "0.1-20" ...
## $ Priority : chr NA NA NA NA ...
## $ Depends : chr "R (>= 1.5.0)" NA "R (>= 3.0.0), car (>= 2.0-19), lmtest, sandwich (>= 2.4-0),\nsurvival (>= 2.37-5), zoo" "R (>= 2.14.0), car, lme4, pbkrtest, reshape2, stringr" ...
## $ Imports : chr "methods, utils" NA "stats, Formula (>= 0.2-0)" NA ...
## $ NeedsCompilation: chr "no" "yes" "no" NA ...
## $ Built : chr "4.0.0" "4.0.0" "4.0.0" "4.0.0" ...A function to do the hard work
As I mentioned above the stack overflow post was a good start but I wanted more
information from the function. Rather than TRUE/FALSE to is it github I would
like as much information as possible about where I got the package. The
package~source function will be applied to the Package column for each row
of our dataframe. For example
as.character(packageDescription("ggplot2")$Repository) will get back “CRAN”,
and as.character(packageDescription("CHAID")$Repository) will yield “R-Forge”.
For GitHub packages the result is character(0) which has a length of zero.
So we’ll test with an if else clause. If we get an answer like “CRAN” we’ll
just return it. If not, we’ll see if there is a GitHub repo listed with
as.character(packageDescription(pkg)$GithubRepo) as well as a GitHub username
as.character(packageDescription(pkg)$GithubUsername). If they exist we’ll
concatenate and return. If not we’ll return “Other”. Besides being good
defensive programming this may catch the package you have built for yourself as
is the case for me.
package_source <- function(pkg){
x <- as.character(packageDescription(pkg)$Repository)
if (length(x)==0) {
y <- as.character(packageDescription(pkg)$GithubRepo)
z <- as.character(packageDescription(pkg)$GithubUsername)
if (length(y)==0) {
return("Other")
} else {
return(str_c("GitHub repo = ", z, "/", y))
}
} else {
return(x)
}
}
# show the first 60 as an example
head(sapply(allmypackages$Package, package_source), 60)## abind acepack AER
## "CRAN" "CRAN" "CRAN"
## afe afex agricolae
## "R-Forge" "CRAN" "CRAN"
## AlgDesign alphavantager Amelia
## "CRAN" "CRAN" "CRAN"
## antiword anytime ape
## "CRAN" "CRAN" "CRAN"
## arm ash askpass
## "CRAN" "CRAN" "CRAN"
## assertthat backports base64enc
## "CRAN" "CRAN" "CRAN"
## BayesFactor bayesplot bayestestR
## "CRAN" "CRAN" "CRAN"
## bbmle BDgraph bdsmatrix
## "CRAN" "CRAN" "CRAN"
## beeswarm BFEffects BH
## "CRAN" "GitHub repo = mattansb/BFEffects" "CRAN"
## bibtex bigmemory bigmemory.sri
## "CRAN" "CRAN" "CRAN"
## bindr bindrcpp binom
## "CRAN" "CRAN" "CRAN"
## BiocManager bit bit64
## "CRAN" "CRAN" "CRAN"
## bitops blavaan blob
## "CRAN" "CRAN" "CRAN"
## blogdown BMA bookdown
## "GitHub repo = rstudio/blogdown" "CRAN" "CRAN"
## Boom BoomSpikeSlab boot
## "CRAN" "CRAN" "CRAN"
## brew bridgesampling brio
## "CRAN" "CRAN" "CRAN"
## brms Brobdingnag broom
## "CRAN" "CRAN" "CRAN"
## broom.mixed broomExtra BSDA
## "CRAN" "CRAN" "CRAN"
## bsts BWStest ca
## "CRAN" "CRAN" "CRAN"
## Cairo callr candisc
## "CRAN" "CRAN" "CRAN"What’s in your libraries?
Now that we have the package_source function we can add a column to our data
frame and do a little looking.
allmypackages$whereat <- sapply(allmypackages$Package, package_source)
str(allmypackages)## 'data.frame': 718 obs. of 9 variables:
## $ Package : chr "abind" "acepack" "AER" "afe" ...
## $ LibPath : chr "/Library/Frameworks/R.framework/Versions/4.0/Resources/library" "/Library/Frameworks/R.framework/Versions/4.0/Resources/library" "/Library/Frameworks/R.framework/Versions/4.0/Resources/library" "/Library/Frameworks/R.framework/Versions/4.0/Resources/library" ...
## $ Version : chr "1.4-5" "1.4.1" "1.2-9" "0.1-20" ...
## $ Priority : chr NA NA NA NA ...
## $ Depends : chr "R (>= 1.5.0)" NA "R (>= 3.0.0), car (>= 2.0-19), lmtest, sandwich (>= 2.4-0),\nsurvival (>= 2.37-5), zoo" "R (>= 2.14.0), car, lme4, pbkrtest, reshape2, stringr" ...
## $ Imports : chr "methods, utils" NA "stats, Formula (>= 0.2-0)" NA ...
## $ NeedsCompilation: chr "no" "yes" "no" NA ...
## $ Built : chr "4.0.0" "4.0.0" "4.0.0" "4.0.0" ...
## $ whereat : chr "CRAN" "CRAN" "CRAN" "R-Forge" ...table(allmypackages$whereat)##
## CRAN GitHub repo = cjtexas/colourgen
## 704 1
## GitHub repo = duncantl/RWordPress GitHub repo = duncantl/XMLRPC
## 1 1
## GitHub repo = easystats/report GitHub repo = haleyjeppson/ggmosaic
## 1 1
## GitHub repo = joachim-gassen/tidycovid19 GitHub repo = leeper/slopegraph
## 1 1
## GitHub repo = lorenzwalthert/stylermd GitHub repo = mattansb/BFEffects
## 1 1
## GitHub repo = PhDMeiwp/Steel.Dwass.test GitHub repo = rstudio/blogdown
## 1 1
## Other R-Forge
## 1 2allmypackages %>%
filter(whereat == "Other") %>%
select(Package, Version)## Package Version
## CGPfunctions CGPfunctions 0.6.3And just to be on the safe side we’ll also write a copy out as a csv file so we have it around in case we ever need to refer back.
write.csv(allmypackages, "mypackagelistMay2019.csv")Go ahead and install R 3.6.0
At this point we have what we need, so go ahead and download and install R 3.6.0. At the end of the installation process you’ll have a pristine copy with a new library directory. When next you restart R and R Studio you’ll see a clean new version. Let’s make use of our data frame to automate most of the process of getting nice clean copies of the libraries we want.
We’ll start by getting the entire tidyverse since we need several parts and
because installing it will trigger the installation of quite a few dependencies
and bootstrap our work.
# post upgrade with output surpessed
install.packages("tidyverse")
library(tidyverse)Now we have R 3.6.0 and some additional packages. Let’s see what we can do.
First let’s create two dataframes, one with our old list and one with what we
have right now. Then we can use anti_join to make a dataframe that lists the
differences thediff. We can use filter and pull to generate a vector of
just the the packages that are on CRAN we want to install.
Note – I’m faking
the output rather than reinstalling all these packages on my machine so you will
see packages from the tidyverse in the listing
oldpackages <- read.csv("mypackagelistMay2019.csv")
allmypackages <- as.data.frame(installed.packages())
allmypackages <- allmypackages %>%
filter(Priority != "base" | is.na(Priority)) %>%
select(-c(Enhances:MD5sum, LinkingTo:Suggests))
thediff <- anti_join(oldpackages,allmypackages, by = "Package")
thediff <- droplevels(thediff)
thediff %>%
filter(whereat == "CRAN") %>%
pull(Package) %>%
as.character ## [1] "abind" "acepack" "AER" "afex" "agricolae" "AlgDesign"
## [7] "alphavantager" "Amelia" "antiword" "anytime" "ape" "arm"
## [13] "ash" "askpass" "assertthat" "backports" "base64enc" "BayesFactor"
## [19] "bayesplot" "bayestestR" "bbmle" "BDgraph" "bdsmatrix" "beeswarm"
## [25] "BH" "bibtex" "bigmemory" "bigmemory.sri" "bindr" "bindrcpp"
## [31] "binom" "BiocManager" "bit" "bit64" "bitops" "blavaan"
## [37] "blob" "BMA" "bookdown" "Boom" "BoomSpikeSlab" "boot"
## [43] "brew" "bridgesampling" "brio" "brms" "Brobdingnag" "broom"
## [49] "broom.mixed" "broomExtra" "BSDA" "bsts" "BWStest" "ca"
## [55] "Cairo" "callr" "candisc" "captioner" "car" "carData"Just do it!
Now that you have a nice automated list of everything that is a CRAN package you can give it a final look and see if there is anything else you’d like to filter out. Once you are sure the list is right one final pipe will set the process in motion.
thediff %>%
filter(whereat == "CRAN") %>%
pull(Package) %>%
as.character %>%
install.packagesDepending on the speed of your network connection and the number of packages you have that will run for a few minutes.
That takes care of our CRAN packages. What about GitHub?
thediff %>%
filter(str_detect(whereat, "GitHub repo")) %>%
select(Package, Version, NeedsCompilation, whereat)## Package Version NeedsCompilation whereat
## 1 BFEffects 0.6.1 no GitHub repo = mattansb/BFEffects
## 2 blogdown 0.21.19 no GitHub repo = rstudio/blogdown
## 3 colourgen 0.2.0 no GitHub repo = cjtexas/colourgen
## 4 ggmosaic 0.3.0 no GitHub repo = haleyjeppson/ggmosaic
## 5 report 0.1.0 no GitHub repo = easystats/report
## 6 RWordPress 0.2-3 no GitHub repo = duncantl/RWordPress
## 7 slopegraph 0.1.14 no GitHub repo = leeper/slopegraph
## 8 Steel.Dwass.test 1.0.0 no GitHub repo = PhDMeiwp/Steel.Dwass.test
## 9 stylermd 0.1.0.9000 no GitHub repo = lorenzwalthert/stylermd
## 10 tidycovid19 0.0.0.9000 no GitHub repo = joachim-gassen/tidycovid19
## 11 XMLRPC 0.3-1 no GitHub repo = duncantl/XMLRPCHere’s another chance to review what you have and whether you still want need
these packages. I could automate the process and once again feed the right
vector to devtools::install_github() but instead I choose to handle these
manually as in devtools::install_github("leeper/slopegraph").
Same with the one package I get from R-Forge…
allmypackages %>%
filter(str_detect(whereat, "R-Forge")) %>%
select(Package, Version, NeedsCompilation, whereat)
install.packages("CHAID", repos="http://R-Forge.R-project.org")At the end of this process you should have a nice clean R install that has all the packages you choose to maintain as well as a detailed listing of what those are.
Done!
I hope you’ve found this useful. I am always open to comments, corrections and suggestions.
Chuck (ibecav at gmail dot com)