Analyst Onboarding

Your setup guide,
start to finish.

Everything you need to go from zero to running analyses — what ARC is, how to get on the network, and how to use RStudio Server.

Welcome Analyst Guide SharePoint Guide
Contents
  1. What is ARC?
  2. Tailscale — Getting on the network
  3. RStudio Server — Setup & environment management
◆

What is ARC?

ARC Institute (Applied Research Center for Health) is an implementation and innovation-focused research organization based in Surabaya, East Java, Indonesia. We do applied health research — not just publishing papers, but making sure the evidence actually reaches policy and practice.

The institute is organized into ten research centers, each covering a specific domain. You'll be working within one or more of these depending on your assignment.

Research Center Structure
Code Focus Area
ARC0Multipurpose / Cross-cutting
ARC1Health Systems & Policy
ARC2Cardiovascular Disease
ARC3Infectious Diseases
ARC4Maternal & Child Health
ARC5Mental Health
ARC6Disability & Rehabilitation
ARC7Nutrition & NCD
ARC8Global Surgery
ARC9Digital Health

Work is tracked in the ARC Project Manager — a dashboard accessible via Tailscale once you're set up. Your lead will share the URL and create your account.

How we work: Projects move through stages — Discussion → Initial Analysis → Initial Result → Manuscript. You'll find your tasks in the project manager dashboard. Update your project's CLAUDE.md checklist as you progress; the dashboard reads it automatically each night.

Tailscale — Getting on the network

ARC's servers are not exposed to the public internet. All internal tools — RStudio, the project manager, data storage — are accessible only through Tailscale, a zero-config mesh VPN. Once you join the ARC tailnet, your device gets a stable private IP and can reach everything directly.

💻
Your Laptop
⟷
🔒
Tailscale
encrypted
⟷
🖥
ARC Server
Before you start: Ask your lead to send you a Tailscale invite to the ARC tailnet. You'll receive an email from Tailscale — you need to accept it before the steps below.
  1. 1
    Download Tailscale for your OS from tailscale.com/download Available for Windows, macOS, Linux, iOS, Android.
  2. 2
    Sign in — open Tailscale and log in with the account you used to accept the invite. You should see the ARC tailnet listed. If you see "no networks," check your email for the invite link first.
  3. 3
    Verify the connection — once connected, open a terminal and ping the server address your lead provides: 
    ping <server-address>
# Should respond — server is reachable
  4. 4
    Keep it running — Tailscale should auto-start with your OS. If you ever lose access to internal tools, check the Tailscale icon in your system tray — it should show "Connected." On mobile: Tailscale must be enabled manually each session unless you set it to auto-connect.
Troubleshooting: If ping fails but Tailscale shows "Connected," the server may be offline — message your lead. If Tailscale itself won't connect, try switching between "Direct" and "Relay" modes in the Tailscale preferences.

RStudio Server — Setup & environment

ARC runs RStudio Server — a full RStudio IDE in your browser, running on the ARC server with all the data already there. No local R install needed.

Requires Tailscale. Complete section 2 first. Your lead will give you the server IP and your login credentials.

Accessing RStudio

  1. 1
    Open your browser while Tailscale is connected and go to: 
    http://<server-ip>:8787
# Replace <server-ip> with the address your lead provides
  2. 2
    Log in with your assigned username and the temporary password. Change your password immediately after first login. RStudio > Tools > Shell, then: passwd
  3. 3
    Open a project — use File > Open Project to navigate to your assigned project folder. Always work inside a project, never from the home directory root.

First-time environment setup

Configure blank slate (do this once) Required

RStudio's default is to save and restore your workspace on every open/close. This causes subtle bugs. Disable it permanently.

Via menu: Tools → Global Options → General → set "Save workspace to .RData on exit" to Never, and uncheck "Restore .RData into workspace at startup."

Or via R console: 

# Run once, then restart
usethis::use_blank_slate()
Restore project packages with renv Required per project

All ARC projects use renv for reproducible package management. Run this every time you open a new project — it installs the exact package versions recorded in the lockfile. 

# In the project console
renv::restore()

# If renv isn't installed yet
install.packages("renv")
renv::restore()
Never run install.packages() without renv active — it installs globally and can break other analysts' environments.

Cleaning your environment

Between analyses or when something behaves unexpectedly, clean your session:

# Remove all objects from memory
rm(list = ls())
gc()             # release memory back to OS

# Full session restart (recommended over rm)
.rs.restartR()   # RStudio only — Ctrl+Shift+F10
Prefer Ctrl+Shift+F10 (restart R) over rm(list=ls()). A restart clears hidden state — loaded packages, options, modified globals — that rm() leaves behind.

Project folder structure

project-name/
├── CLAUDE.md        # checklist — update as you progress stages
├── data/            # raw input files (never commit)
├── scripts/         # R scripts, numbered in run order
│   ├── 01_clean.R
│   ├── 02_analysis.R
│   └── 03_figures.R
├── output/          # figures, tables
├── manuscript/      # draft documents
└── renv/            # package lockfile — always commit this

Git basics

# Never commit raw data or large files
echo "data/" >> .gitignore

# Commit often, with meaningful messages
git add scripts/ output/ renv.lock CLAUDE.md
git commit -m "complete variable harmonization step"
git push

Work on a feature branch, not directly on main:

git checkout -b analysis/your-name

© ARC Institute

← Welcome SharePoint Guide →