R Scripts and Projects

Introduction

As you progress from simple calculations to complex analyses, organizing your R work becomes crucial. R scripts and RStudio Projects are the foundation of reproducible, organized data analysis. In this lesson, you’ll learn how to structure your R code professionally and manage your analytical projects effectively.

Think of R scripts as recipes and Projects as well-organized kitchens – both are essential for creating consistent, reproducible results.

Understanding R Scripts

What is an R Script?

An R script is a plain text file with a .R extension that contains R commands. Unlike typing commands directly in the console, scripts allow you to:

  • Save your work for future use
  • Document your analysis with comments
  • Share your methods with others
  • Reproduce results exactly
  • Debug and modify code systematically

Creating Your First Script

  1. File > New File > R Script (or Ctrl+Shift+N)
  2. Save immediately: File > Save (or Ctrl+S)
  3. Give it a meaningful name: my_first_analysis.R

Script Structure Best Practices

Here’s a well-structured R script template:

# ============================================================================
# Project: Introduction to R Analysis
# Script: my_first_analysis.R
# Author: Your Name
# Date: 2024-09-21
# Last Modified: 2024-09-21
#
# Purpose: Demonstrate best practices for R script organization
#
# Data: Built-in R datasets
# Output: Summary statistics and basic plots
# ============================================================================

# SETUP ======================================================================

# Clear workspace (optional, be careful!)
# rm(list = ls())

# Load required libraries
library(tidyverse)  # Data manipulation and visualization
library(here)       # File path management

# Set working directory (if not using projects)
# setwd("~/Documents/R-Analysis")

# Source additional scripts if needed
# source("helper_functions.R")

# CONSTANTS AND CONFIGURATION ================================================

# Define constants
SIGNIFICANCE_LEVEL <- 0.05
DEFAULT_PLOT_WIDTH <- 8
DEFAULT_PLOT_HEIGHT <- 6

# Set plot theme
theme_set(theme_minimal())

# DATA IMPORT ================================================================

# Load built-in dataset for this example
data("mtcars")

# In real projects, you might load data like this:
# my_data <- read_csv(here("data", "raw", "dataset.csv"))

# DATA EXPLORATION ===========================================================

# Quick overview
glimpse(mtcars)
summary(mtcars)

# Check for missing values
sum(is.na(mtcars))

# DATA ANALYSIS ==============================================================

# Calculate summary statistics
mpg_stats <- mtcars %>%
  summarise(
    mean_mpg = mean(mpg),
    median_mpg = median(mpg),
    sd_mpg = sd(mpg),
    min_mpg = min(mpg),
    max_mpg = max(mpg)
  )

print(mpg_stats)

# Create visualizations
mpg_histogram <- ggplot(mtcars, aes(x = mpg)) +
  geom_histogram(bins = 10, fill = "steelblue", alpha = 0.7) +
  labs(
    title = "Distribution of Miles Per Gallon",
    x = "Miles Per Gallon",
    y = "Frequency"
  )

print(mpg_histogram)

# RESULTS EXPORT =============================================================

# Save plot
ggsave(
  filename = here("output", "mpg_distribution.png"),
  plot = mpg_histogram,
  width = DEFAULT_PLOT_WIDTH,
  height = DEFAULT_PLOT_HEIGHT,
  dpi = 300
)

# Save results
write_csv(mpg_stats, here("output", "mpg_summary_stats.csv"))

# SESSION INFO ===============================================================

# Document session for reproducibility
sessionInfo()

# End of script

Key Script Elements

1. Header Comments

Always start with a comprehensive header: - Project name and purpose - Author and date information - Brief description of what the script does - Data sources and outputs

2. Setup Section

  • Load required libraries
  • Set global options
  • Define constants
  • Source helper functions

3. Organized Sections

Use clear section headers:

# SECTION NAME ================================================================
# or
# Section Name ----

4. Meaningful Comments

# Good comments explain WHY, not just WHAT
mpg_threshold <- 20  # Fuel efficiency benchmark for classification

# Bad comment (obvious)
mpg_threshold <- 20  # Set mpg_threshold to 20

Working with Scripts

Running Script Code

Option 1: Line by Line - Place cursor on line - Press Ctrl+Enter (Windows/Linux) or Cmd+Enter (Mac)

Option 2: Selected Code - Highlight code block - Press Ctrl+Enter

Option 3: Entire Script - Press Ctrl+Shift+Enter or click “Source”

Option 4: From Console

source("my_script.R")

Script Navigation

Code Sections Create collapsible sections with ---- or ====:

# Data Import ----
# (code here)

# Data Cleaning ====
# (code here)

Go to Function/Variable - Ctrl+. opens “Go to File/Function” dialog - Type name to jump quickly

Outline View - Click the outline button in Source pane - See all functions and sections

Introduction to RStudio Projects

Why Use Projects?

RStudio Projects solve common organizational problems:

Without Projects:

# Absolute paths (bad!)
setwd("/Users/john/Documents/my_analysis")
data <- read.csv("/Users/john/Documents/my_analysis/data/file.csv")

# Problems:
# - Not portable between computers
# - Difficult to share
# - Hard to organize multiple analyses

With Projects:

# Relative paths (good!)
data <- read.csv("data/file.csv")
# or even better:
data <- read_csv(here("data", "file.csv"))

Project Benefits

  1. Automatic working directory - no more setwd()
  2. Organized file structure - everything in one place
  3. Portable - works on any computer
  4. Version control ready - easy Git integration
  5. Workspace isolation - separate environments for different projects

Creating a New Project

Method 1: From RStudio 1. File > New Project 2. Choose project type: - New Directory: Start fresh - Existing Directory: Use existing folder - Version Control: Clone from Git

Method 2: New Directory Walkthrough 1. Select “New Directory” 2. Choose “New Project” 3. Directory name: my-r-analysis 4. Choose parent directory 5. Optional: Initialize Git repository 6. Click “Create Project”

Setting Up Project Structure

Create this structure using R:

# Create project folders
dir.create("data")
dir.create("data/raw")
dir.create("data/processed")
dir.create("scripts")
dir.create("output")
dir.create("output/figures")
dir.create("output/tables")
dir.create("docs")

# Create README file
writeLines(
  c("# My R Analysis Project",
    "",
    "## Description",
    "Brief description of your project",
    "",
    "## Structure",
    "- `data/`: Data files",
    "- `scripts/`: R scripts",
    "- `output/`: Results and figures",
    "- `docs/`: Documentation"),
  "README.md"
)

File Management Best Practices

Naming Conventions

Files and Folders: - Use lowercase with hyphens: data-cleaning.R - Or use underscores: data_cleaning.R - Be descriptive: monthly_sales_analysis.R - Use prefixes for ordering: 01_import.R, 02_clean.R

Variables and Functions:

# Good naming
customer_ages <- c(25, 30, 45, 22)
calculate_total_revenue <- function(price, quantity) { ... }

# Avoid
x <- c(25, 30, 45, 22)
fun1 <- function(a, b) { ... }

Working with Paths

Use the here Package:

# Install if needed
install.packages("here")
library(here)

# Benefits of here()
data_path <- here("data", "raw", "sales.csv")
# Works on Windows: data/raw/sales.csv
# Works on Mac/Linux: data/raw/sales.csv

# Read data
sales_data <- read_csv(here("data", "raw", "sales.csv"))

# Save output
write_csv(results, here("output", "tables", "summary.csv"))

Data Import Best Practices

# Good data import workflow
import_sales_data <- function() {
  # Define file path
  file_path <- here("data", "raw", "sales_2024.csv")

  # Check if file exists
  if (!file.exists(file_path)) {
    stop("Sales data file not found: ", file_path)
  }

  # Import with explicit column types
  sales_data <- read_csv(
    file_path,
    col_types = cols(
      date = col_date(format = "%Y-%m-%d"),
      customer_id = col_character(),
      amount = col_double(),
      product = col_character()
    )
  )

  # Basic validation
  if (nrow(sales_data) == 0) {
    warning("Sales data is empty")
  }

  return(sales_data)
}

# Use the function
sales_data <- import_sales_data()

Reproducible Analysis Workflow

Script Dependencies

Master Script Approach:

# main_analysis.R
# ============================================================================
# Master Script: Sales Analysis Pipeline
# ============================================================================

# Setup
source(here("scripts", "00_setup.R"))

# Data pipeline
source(here("scripts", "01_data_import.R"))
source(here("scripts", "02_data_cleaning.R"))
source(here("scripts", "03_data_analysis.R"))
source(here("scripts", "04_generate_report.R"))

# Session info
sessionInfo()

Package Management

Using renv for Reproducibility:

# Install renv
install.packages("renv")

# Initialize project environment
renv::init()

# Install packages
install.packages(c("tidyverse", "here", "lubridate"))

# Take snapshot of current packages
renv::snapshot()

# Restore environment (on another computer)
renv::restore()

Common Project Pitfalls

Pitfall 1: Hardcoded Paths

# Bad
setwd("C:/Users/John/Documents/analysis")
data <- read.csv("C:/Users/John/Documents/analysis/data.csv")

# Good
data <- read_csv(here("data", "data.csv"))

Pitfall 2: Not Using Version Control

  • Initialize Git repository when creating project
  • Use meaningful commit messages
  • Push to GitHub/GitLab for backup

Pitfall 3: Mixing Data and Scripts

# Bad structure
project/
├── analysis.R
├── data.csv
├── plot.R
├── results.csv

# Good structure
project/
├── scripts/
│   ├── analysis.R
│   └── plot.R
├── data/
│   └── data.csv
└── output/
    └── results.csv

Pitfall 4: No Documentation

Always include: - README.md with project description - Comments in scripts explaining complex logic - Data dictionaries for datasets

Practical Exercise

Let’s create a complete project:

  1. Create New Project

    • Name: “iris-analysis”
    • Choose appropriate location

  2. Set Up Structure

    # Run this in your new project
dir.create("data")
dir.create("scripts")
dir.create("output")

  3. Create Analysis Script

    # scripts/iris_analysis.R

# ========================================================================
# Iris Dataset Analysis
# Author: Your Name
# Date: Today's Date
# ========================================================================

# Setup
library(tidyverse)
library(here)

# Load data (built-in dataset)
data("iris")

# Quick exploration
glimpse(iris)
summary(iris)

# Analysis
species_summary <- iris %>%
  group_by(Species) %>%
  summarise(
    mean_sepal_length = mean(Sepal.Length),
    mean_petal_length = mean(Petal.Length),
    count = n()
  )

print(species_summary)

# Visualization
scatter_plot <- ggplot(iris, aes(x = Sepal.Length, y = Petal.Length, color = Species)) +
  geom_point(size = 3) +
  labs(
    title = "Iris: Sepal Length vs Petal Length",
    x = "Sepal Length (cm)",
    y = "Petal Length (cm)"
  ) +
  theme_minimal()

print(scatter_plot)

# Save results
write_csv(species_summary, here("output", "species_summary.csv"))
ggsave(here("output", "iris_scatter.png"), scatter_plot, width = 8, height = 6)

  4. Run and Test

    • Run script section by section
    • Check that files are created in output folder
    • Verify paths work correctly

Summary

You now understand how to organize R work professionally:

R Scripts

  • Structure scripts with clear headers and sections
  • Comment code to explain your reasoning
  • Use meaningful variable and file names
  • Organize code logically from setup to results

RStudio Projects

  • Create projects for each analysis
  • Use consistent folder structures
  • Leverage relative paths with here()
  • Document your work with README files

Best Practices

  • Plan your analysis structure before coding
  • Test scripts section by section
  • Save intermediate results
  • Document your session info for reproducibility

Next Steps

With Module 1 complete, you’re ready to dive deeper into R:

Pro Tips for Success
  1. Start every analysis with a new project
  2. Use the same folder structure across projects
  3. Write scripts as if someone else will read them
  4. Save your work frequently and use version control
  5. Test code in small chunks before running entire scripts

Congratulations! You’ve completed Module 1 and have a solid foundation for R programming. The habits you develop now will serve you well throughout your data science journey.