Before Day 1

Software setup — please complete this before you arrive

Budget 45 minutes. Everything below is free. If something fails, bring the error message to the first session rather than skipping the step.

1. Install R (≥ 4.3)

Download from CRAN. Take the latest release. R 4.3 is the floor because the workshop uses the native pipe |> and dplyr 1.1 syntax throughout.

Check inside R:

getRversion()

2. Install an IDE

Positron is what the facilitator will be using — it handles both R and Python and has a built-in AI assistant we use on Day 5. RStudio works for everything except the Day 5 assistant demo.

3. Install Quarto (≥ 1.4)

Download from quarto.org. Positron and recent RStudio bundle a copy, but a system install keeps the command line working:

quarto check

4. Get the project

git clone https://github.com/Zahedasghar/d4d.git

No Git yet? Download the ZIP from the repository page — but note that Day 1 covers Git, so installing it now saves time later.

Open the folder by double-clicking d4d.Rproj. This matters: it sets the project root that here::here() resolves against, and every script in this repository depends on that.

5. Install the R packages

In the R console, from inside the project:

source("R/setup-packages.R")

This installs roughly 35 packages and takes 10–20 minutes on a first run. It stops with an explicit error listing anything that failed, rather than continuing quietly.

If the spatial packages fail

sf, tmap and leaflet need system libraries (GDAL, GEOS, PROJ). On Windows and macOS the installer usually handles this; on Linux you may need:

sudo apt install libgdal-dev libgeos-dev libproj-dev libudunits2-dev

These are only needed for the Day 3 afternoon. Everything else will run without them.

6. Check the data pack

The data/ folder should contain the survey files and Pakistan shapefiles. Confirm:

list.files(here::here("data"))

If data/ is empty or missing the large .DTA and .sav files, ask the facilitator — some files are distributed separately because of their size.

7. Verify

Run this. All four lines should print TRUE.

library(tidyverse)
library(here)

c(
  r_version_ok = getRversion() >= "4.3.0",
  here_finds_root = file.exists(here::here("d4d.Rproj")),
  data_present = dir.exists(here::here("data")),
  quarto_found = nzchar(Sys.which("quarto"))
)

Anything FALSE — bring it to the first session.