Setup R-GeoSpatial Environment for HPC

This repository contains a Shell script (setup.sh) to set up an R environment with necessary dependencies and configurations for geocomputation on an Ubuntu system HPC.

Prerequisites

• Ubuntu: The script is designed for Ubuntu systems. Ensure you are running a compatible version.
• sudo privileges: You need administrative privileges to install packages and modify system settings.

Steps to Download and Run the Script

1. Download the Script

You can download the setup.sh script directly from GitHub using the following command:

wget https://raw.githubusercontent.com/ai4city-hkust/RSpatialHPC/main/setup.sh

This will download the script to your current directory.

2. Make the Script Executable

Before running the script, you need to make it executable. Run the following command:

chmod +x setup.sh

This grants execution permissions to the script.

3. Run the Script

Execute the script with the following command:

./setup.sh

The script will:

1. Prompt you for your sudo password to cache it for the session.
2. Install necessary system dependencies.
3. Set up R and configure the R package installation path.
4. Install required R packages.
5. Optionally install RStudio Server (if you choose to).

4. Follow On-Screen Instructions

The script is interactive and will guide you through the process. For example:

• You will be asked if you want to install RStudio Server.
• The script will automatically configure the R library paths based on your system.

5. Verify the Setup

After the script completes, you can verify the setup by starting R and checking the library paths:

R

Inside the R console, run:

.libPaths()

This will display the configured library paths, ensuring the setup was successful.

---

Script Details

The script performs the following tasks:

1. System Information: Displays Ubuntu version and CPU information.
2. System Updates: Updates the package indices and installs necessary helper packages.
3. R Installation: Adds the CRAN repository and installs base R.
4. R Dependencies: Installs system dependencies required for R packages.
5. RStudio Server (Optional): Installs RStudio Server if requested.
6. R Library Path Configuration: Sets up a custom R library path (~/pkgR or ~/pkgRUserLocal) and writes it to ~/.Rprofile for persistence.
7. R Package Installation: Installs essential R packages like tidyverse, sf, terra, and others.

---

Notes

• RStudio Server: Installation of RStudio Server is optional. If you choose not to install it, you can still use R from the terminal or install RStudio Desktop separately.
• Custom Library Path: The script checks if ~/pkgR exists. If it does, it uses ~/pkgRUserLocal as the R library path; otherwise, it uses ~/pkgR.
• Persistent Configuration: The R library path is written to ~/.Rprofile, ensuring it is applied every time R starts.

---

Troubleshooting

• Permission Issues: Ensure you have sudo privileges to run the script.
• Network Issues: If the script fails to download packages or keys, check your internet connection.
• R Path Issues: If .libPaths() does not show the expected paths, manually check the ~/.Rprofile file for correct configuration.