Getting to know your R Setup

class: center, middle, inverse, title-slide

# Getting to know your R Setup
## ⚔ and other demystifications
### Julia Piaskowski
### 2022-03-03

---

## Updating R
*(please don't do this in the middle of the workshop)*

**Windows**
```
library(installr); updateR() # follow prompts
```

**Mac**  Visit the [website](https://cran.r-project.org/bin/macosx/), download and install.

**Linux**   Run in terminal: 
```
sudo apt update
sudo apt upgrade
```

---
## Updating packages

* Command line - single package:

```r
devtools::update.packages("RColorBrewer")
```

* Command line - all packages:

*(don't run this unless you have the time to wait)*

```r
devtools::update_packages(TRUE)
```

* point-and-click using RStudio's "Packages" Pane

---
## Downgrade a package

Easiest way (if you know the version):

```r
devtools::install_version("RColorBrewer", "1.0.5")
```

Downgrade to a specific point in time:   
*(note that going this far back in time usually causes problems)*

```r
groundhog::groundhog.library("RColorBrewer", "2010-01-01")
```
How to find the version information? Visit the [package site](http://localhost:4321/2022-03-01-WhatTheyForgot) on [CRAN](https://cran.r-project.org/web/packages/available_packages_by_name.html). 
---
## Installing packages when you update R

* If doing a minor update of R (e.g. 4.0.1 to 4.0.3), do nothing (packages will remain compatible)
* If doing a major R update (4.0 to 4.1), re-install packages. *Don't copy over packages directly because things may break without telling you.*

One possible [solution](https://rstats.wtf/maintaining-r.html):

```r
# Install new version of R (4.1 in this example) Create a
# new directory for the version of R
fs::dir_create("~/Library/R/4.1/library")
# Re-start R so the .libPaths are updated OR find where
# your R library exists already with .libPaths() Look up
# what packages were in your old package library
pkgs <- fs::path_file(fs::dir_ls("~/Library/R/3.6/library"))
# Filter these packages as needed Install the packages in
# the new version
install.packages(pkgs)
```

---
## Where does R live?

---
## Where the R executable(s) live(s)

This tell us the path to R that RStudio is using:

```r
R.home()
```

```
## [1] "/usr/lib/R"
```

It can sometimes be more [complicated](https://support.rstudio.com/hc/en-us/articles/200486138-Changing-R-versions-for-the-RStudio-Desktop-IDE) than that depending on the system.

---
## Where your package libraries live

```r
.libPaths()
```

```
## [1] "/home/jpiaskowski/R/x86_64-pc-linux-gnu-library/4.1"
## [2] "/usr/local/lib/R/site-library"                      
## [3] "/usr/lib/R/site-library"                            
## [4] "/usr/lib/R/library"
```
1. main library
1. empty
1. added via the command line and updates with `sudo apt-get update`
1. "base R packages" (what is shipped with R)

R is using all of these, in order.

---
## Check your default library

(by installing a package)

```r
install.packages("RColorBrewer")
```
```
Installing package into 
‘/home/jpiaskowski/R/x86_64-pc-linux-gnu-library/4.1’
(as ‘lib’ is unspecified)
```

---
class: middle, center, inverse
# terminal fun

## [exercise](exercises/exercise-1.html)

---
background-image: url(https://raw.githubusercontent.com/imci-idaho/2022-03-01-WhatTheyForgot/main/static/slides/images/R-startup.svg)
background-position: center
background-size: contain
class: bottom, left

## What Happens When [R is Started?](https://stat.ethz.ch/R-manual/R-devel/library/base/html/Startup.html)

---
#  .Renviron

* Sets global variables: API and authentication keys
* There is often only one of these per R system, located in the home directory, but there can one per project
* DO NOT add this to a git repository or make sure it ends up in  *.gitignore*

--
Example:

```r
if (!file.exists(".Renviron")) file.create(".Renviron")
```

Open file and manually add:

```r
NOAA_KEY = "my-secret-key"
```

Some helper functions from **usethis**:

```r
usethis::edit_r_environ("user")  # system-wide
usethis::edit_r_environ("project")  # for the current project you are in
```

---
#  .Rprofile

* An R script that runs every time R is opened or restarted
* there is (often) one created automatically when R is installed at the home directory
* There may be one in each **.Rproj**, this takes precedence over others
* This often goes in the git repo

---
## A possible .Rprofile

```
options(scipen=999) # force R to not use scientific notation

.First <- function() {
 cat("Welcome, High Priestess of R. 
 You are a true hero and your legend will never die.")
}

library(tidyverse) # probably not a good idea 
source("scripts/file_import.R")

.Last <- function() {
 cat('Praised be, my R programmer.')
}
```

more [details](https://sodocumentation.net/r/topic/4166/-rprofile) and  [options](https://stat.ethz.ch/R-manual/R-devel/library/base/html/options.html)

---
class: center, middle, inverse

.pull-left[
<img src="images/prince.jpg" width="200%" />
]

.pull-right[
[exercise](exercises/exercise-2.html)
]

---
# Namespace conflicts

```r
library(dplyr)
```

```
## 
## Attaching package: 'dplyr'
```

```
## The following objects are masked from 'package:stats':
## 
##     filter, lag
```

```
## The following objects are masked from 'package:base':
## 
##     intersect, setdiff, setequal, union
```

Only one function name can be loaded into an R session. Identical function names for packages loaded earlier must be referenced like thus: `base::filter()`.

---
# Find conflicts:

```r
getAnywhere("filter")
```

```
## 2 differing objects matching 'filter' were found
## in the following places
##   package:dplyr
##   package:stats
##   namespace:dplyr
##   namespace:stats
## Use [] to view one of them
```

---
## Namespace solutions

* Use `::`

```r
dplyr::filter(...)
```

* Remove conflicting packages:

```r
unloadNamespace("offending_package")
```

* Only require packages you will use (reconsider `library(tidyverse)`)
* Load the packages with whose functions you do not want masked last
* Use the package **conflicted**

---
## A namespace solution under development:

```r
# devtools::install_github('r-lib/conflicted')
library(conflicted)
library(dplyr)
conflict_scout()
```

```
## 2 conflicts:
## * `filter`: dplyr, stats
## * `lag`   : dplyr, stats
```

```r
conflict_prefer("filter", "dplyr")
```

```
## [conflicted] Will prefer dplyr::filter over any other package
```

---
# Find all the loaded namespaces:

```r
sort(loadedNamespaces())
```

```
##  [1] "assertthat" "base"       "bslib"      "cachem"     "cli"       
##  [6] "compiler"   "conflicted" "crayon"     "datasets"   "DBI"       
## [11] "digest"     "dplyr"      "ellipsis"   "evaluate"   "fansi"     
## [16] "fastmap"    "formatR"    "generics"   "glue"       "graphics"  
## [21] "grDevices"  "highr"      "htmltools"  "jquerylib"  "jsonlite"  
## [26] "knitr"      "lifecycle"  "magrittr"   "memoise"    "methods"   
## [31] "pillar"     "pkgconfig"  "purrr"      "R6"         "rlang"     
## [36] "rmarkdown"  "rstudioapi" "sass"       "stats"      "stringi"   
## [41] "stringr"    "tibble"     "tidyselect" "tools"      "utf8"      
## [46] "utils"      "vctrs"      "xaringan"   "xfun"       "yaml"
```