# Installing R, IDEs, Packages and Projects — R Programming

Source: https://www.geekswithgeeks.com/en/r-programming/f-setup

> Set up R with an IDE, install packages and keep projects reproducible.

## A productive environment

Install **R** itself from CRAN, then an editor. **RStudio** (from Posit) has long been the standard IDE, with a console, script editor, environment viewer, plots pane and help; **Positron**, Posit's newer data-science IDE built on VS Code foundations, supports both R and Python; and **VS Code** with the R extension also works. Install packages from CRAN with **`install.packages("dplyr")`** (once per machine) and load them in each session with **`library(dplyr)`**. The **tidyverse** meta-package installs and loads core packages together: ggplot2, dplyr, tidyr, readr, purrr, tibble, stringr, forcats and lubridate. Work in **projects** (an `.Rproj` file or folder) so file paths are relative to the project, and avoid `setwd()` with absolute paths that break on other machines. For reproducibility, **`renv`** records exact package versions in a lockfile (`renv::init()`, `renv::snapshot()`, `renv::restore()`), similar to lock files in other languages. Get help with `?mean`, `help(package = "dplyr")` and the vignettes that many packages include.

## Packages and a reproducible project setup

Install once, load every session, and lock versions with renv.

```r
install.packages(c("tidyverse", "renv"))   # once per machine

library(tidyverse)                         # every session

# inside an RStudio/Positron project folder:
renv::init()        # create a project library and renv.lock
# ... install or update packages as you work ...
renv::snapshot()    # record exact versions in renv.lock
# a colleague later runs:
renv::restore()     # installs the same versions

?read_csv                                   # help page
vignette("dplyr")                           # long-form guide
sessionInfo()                               # R and package versions for bug reports
```

## A kitchen with labelled jars

install.packages buys ingredients for your pantry; library() takes the jar off the shelf for today's cooking. renv writes down exactly which brand and batch you used, so someone else can cook the same dish.

**Quiz:** What is the difference between install.packages() and library()?

- [ ] There is none
- [ ] library downloads; install.packages loads
- [ ] Both must be run before every line
- [x] install.packages downloads a package once; library loads it into the current session

*Answer:* install.packages downloads a package once; library loads it into the current session. Installation happens once per machine; loading happens in each R session.