R Programming Language

What is an R file?

An R (.r or .R) file is a source code file written in the R programming language, a powerful statistical computing and graphics language widely used for data analysis, statistical modeling, and data visualization. R files contain scripts, functions, and data manipulation code that can be executed in the R environment to perform complex statistical analyses, create sophisticated visualizations, and develop statistical models and machine learning algorithms.

More Information

R was developed by Ross Ihaka and Robert Gentleman at the University of Auckland, New Zealand, starting in 1993, and was first released to the public in 1995. R is an implementation and extension of the S programming language created at Bell Labs. The language was designed specifically for statistical computing and has become the de facto standard for statistical analysis in academia, research, and increasingly in industry.

R's popularity has grown exponentially due to its comprehensive statistical capabilities, extensive package ecosystem (CRAN), and strong community support. The language excels in data manipulation, statistical modeling, machine learning, and creating publication-quality graphics. R has played a crucial role in the data science revolution and remains one of the most important tools for statisticians, data scientists, researchers, and analysts worldwide.

R Format

R has a flexible syntax designed for interactive data analysis and statistical computing:

Basic Syntax

• Assignment operators - <- (preferred) or = for variable assignment
• Vectorized operations - Operations work on entire vectors
• Function calls - function_name(arguments)
• Comments - Lines starting with # symbol
• Objects - Everything in R is an object
• Case sensitivity - Variable and function names are case-sensitive

Key Features

• Vectorized computing - Efficient operations on entire data structures
• Functional programming - Functions as first-class objects
• Statistical functions - Comprehensive built-in statistical capabilities
• Data frames - Powerful data structure for mixed-type data
• Package system - Extensive library ecosystem
• Interactive environment - REPL for exploratory data analysis

Data Types and Structures

• Basic types - numeric, integer, character, logical, complex
• Vectors - Homogeneous collections of data
• Lists - Heterogeneous collections of objects
• Matrices - Two-dimensional arrays
• Data frames - Tabular data structure (like Excel sheets)
• Factors - Categorical variables
• Arrays - Multi-dimensional data structures

Example R Script

# =====================================================
# Comprehensive Data Analysis Script
# Purpose: Exploratory Data Analysis and Modeling
# Author: Data Scientist
# Date: Sys.Date()
# =====================================================

# Load required libraries
library(tidyverse)    # Data manipulation and visualization
library(ggplot2)      # Advanced plotting
library(dplyr)        # Data manipulation
library(readr)        # Reading data files
library(corrplot)     # Correlation visualization
library(randomForest) # Machine learning
library(caret)        # Classification and regression training
library(plotly)       # Interactive plots

# Set global options
options(scipen = 999)  # Disable scientific notation
set.seed(123)          # For reproducibility

# Create sample dataset (normally you'd read from file)
# Example: data <- read_csv("data.csv")
create_sample_data <- function(n = 1000) {
  data.frame(
    id = 1:n,
    age = sample(18:80, n, replace = TRUE),
    income = rnorm(n, mean = 50000, sd = 15000),
    education = sample(c("High School", "Bachelor's", "Master's", "PhD"), 
                      n, replace = TRUE, prob = c(0.4, 0.35, 0.2, 0.05)),
    experience = pmax(0, rnorm(n, mean = 10, sd = 5)),
    satisfaction = sample(1:10, n, replace = TRUE),
    department = sample(c("Sales", "Engineering", "Marketing", "HR"), 
                       n, replace = TRUE),
    performance = rnorm(n, mean = 7.5, sd = 1.5)
  ) %>%
    mutate(
      # Create derived variables
      income = pmax(25000, income),  # Minimum income
      age_group = cut(age, breaks = c(0, 30, 50, 100), 
                     labels = c("Young", "Middle", "Senior")),
      high_performer = performance > median(performance),
      income_level = cut(income, breaks = quantile(income, probs = 0:4/4),
                        labels = c("Low", "Medium", "High", "Very High"),
                        include.lowest = TRUE)
    )
}

# Load and prepare data
cat("Loading and preparing data...\n")
data <- create_sample_data(1000)

# Data exploration and summary
cat("\n=== DATA SUMMARY ===\n")
str(data)
summary(data)

# Check for missing values
missing_data <- data %>%
  summarise_all(~sum(is.na(.))) %>%
  gather(variable, missing_count) %>%
  filter(missing_count > 0)

if(nrow(missing_data) > 0) {
  print("Missing data found:")
  print(missing_data)
} else {
  cat("No missing data found.\n")
}

# Descriptive statistics function
describe_numeric <- function(x) {
  if(is.numeric(x)) {
    list(
      mean = round(mean(x, na.rm = TRUE), 2),
      median = round(median(x, na.rm = TRUE), 2),
      sd = round(sd(x, na.rm = TRUE), 2),
      min = round(min(x, na.rm = TRUE), 2),
      max = round(max(x, na.rm = TRUE), 2),
      q25 = round(quantile(x, 0.25, na.rm = TRUE), 2),
      q75 = round(quantile(x, 0.75, na.rm = TRUE), 2)
    )
  }
}

# Apply descriptive statistics to numeric columns
numeric_summary <- data %>%
  select_if(is.numeric) %>%
  map(describe_numeric) %>%
  bind_rows(.id = "variable")

print(numeric_summary)

# Visualization section
cat("\n=== CREATING VISUALIZATIONS ===\n")

# 1. Distribution plots
p1 <- ggplot(data, aes(x = income)) +
  geom_histogram(bins = 30, fill = "steelblue", alpha = 0.7) +
  geom_vline(aes(xintercept = mean(income)), color = "red", linetype = "dashed") +
  labs(title = "Income Distribution", 
       subtitle = paste("Mean income: $", round(mean(data$income), 0)),
       x = "Income ($)", y = "Frequency") +
  theme_minimal() +
  scale_x_continuous(labels = scales::dollar_format())

# 2. Box plots by category
p2 <- ggplot(data, aes(x = department, y = performance, fill = department)) +
  geom_boxplot(alpha = 0.7) +
  geom_jitter(width = 0.2, alpha = 0.3) +
  labs(title = "Performance by Department",
       x = "Department", y = "Performance Score") +
  theme_minimal() +
  theme(legend.position = "none")

# 3. Correlation matrix
numeric_data <- data %>% select_if(is.numeric) %>% select(-id)
correlation_matrix <- cor(numeric_data, use = "complete.obs")

# 4. Scatter plot with trend line
p3 <- ggplot(data, aes(x = experience, y = income)) +
  geom_point(aes(color = education), alpha = 0.6) +
  geom_smooth(method = "lm", se = TRUE, color = "black") +
  labs(title = "Income vs Experience by Education Level",
       x = "Years of Experience", y = "Income ($)") +
  scale_y_continuous(labels = scales::dollar_format()) +
  theme_minimal() +
  facet_wrap(~education)

# Display correlation plot
corrplot(correlation_matrix, method = "circle", type = "upper",
         order = "hclust", tl.cex = 0.8, tl.col = "black")

# Statistical analysis
cat("\n=== STATISTICAL ANALYSIS ===\n")

# 1. T-test: Compare performance between high and low income groups
income_median <- median(data$income)
high_income <- data$performance[data$income > income_median]
low_income <- data$performance[data$income <= income_median]

t_test_result <- t.test(high_income, low_income)
cat("T-test: Performance difference between income groups\n")
print(t_test_result)

# 2. ANOVA: Performance differences across departments
anova_result <- aov(performance ~ department, data = data)
cat("\nANOVA: Performance differences across departments\n")
print(summary(anova_result))

# Post-hoc test if ANOVA is significant
if(summary(anova_result)[[1]][["Pr(>F)"]][1] < 0.05) {
  tukey_result <- TukeyHSD(anova_result)
  print(tukey_result)
}

# 3. Linear regression model
cat("\n=== PREDICTIVE MODELING ===\n")

# Prepare data for modeling
model_data <- data %>%
  select(performance, age, income, experience, satisfaction) %>%
  na.omit()

# Split data into training and testing sets
train_index <- createDataPartition(model_data$performance, p = 0.7, list = FALSE)
train_data <- model_data[train_index, ]
test_data <- model_data[-train_index, ]

# Linear regression model
lm_model <- lm(performance ~ age + income + experience + satisfaction, 
               data = train_data)

cat("Linear Regression Model Summary:\n")
print(summary(lm_model))

# Model predictions
predictions <- predict(lm_model, test_data)
rmse <- sqrt(mean((test_data$performance - predictions)^2))
r_squared <- cor(test_data$performance, predictions)^2

cat(sprintf("\nModel Performance:\n"))
cat(sprintf("RMSE: %.3f\n", rmse))
cat(sprintf("R-squared: %.3f\n", r_squared))

# Random Forest model for comparison
rf_model <- randomForest(performance ~ age + income + experience + satisfaction,
                        data = train_data, ntree = 100)

rf_predictions <- predict(rf_model, test_data)
rf_rmse <- sqrt(mean((test_data$performance - rf_predictions)^2))
rf_r_squared <- cor(test_data$performance, rf_predictions)^2

cat(sprintf("\nRandom Forest Performance:\n"))
cat(sprintf("RMSE: %.3f\n", rf_rmse))
cat(sprintf("R-squared: %.3f\n", rf_r_squared))

# Feature importance from Random Forest
importance_data <- data.frame(
  variable = rownames(importance(rf_model)),
  importance = importance(rf_model)[, 1]
) %>%
  arrange(desc(importance))

print("Feature Importance (Random Forest):")
print(importance_data)

# Export results
cat("\n=== EXPORTING RESULTS ===\n")

# Create results summary
results_summary <- list(
  data_summary = list(
    n_observations = nrow(data),
    n_variables = ncol(data),
    numeric_summary = numeric_summary
  ),
  statistical_tests = list(
    t_test_p_value = t_test_result$p.value,
    anova_p_value = summary(anova_result)[[1]][["Pr(>F)"]][1]
  ),
  model_performance = list(
    linear_regression = list(rmse = rmse, r_squared = r_squared),
    random_forest = list(rmse = rf_rmse, r_squared = rf_r_squared)
  ),
  feature_importance = importance_data
)

# Save plots
ggsave("income_distribution.png", p1, width = 10, height = 6, dpi = 300)
ggsave("performance_by_department.png", p2, width = 10, height = 6, dpi = 300)
ggsave("income_vs_experience.png", p3, width = 12, height = 8, dpi = 300)

# Save data and results
write_csv(data, "processed_data.csv")
saveRDS(results_summary, "analysis_results.rds")
saveRDS(lm_model, "linear_model.rds")
saveRDS(rf_model, "random_forest_model.rds")

cat("Analysis complete! Results and plots saved.\n")
cat("Files created:\n")
cat("- processed_data.csv\n")
cat("- analysis_results.rds\n")
cat("- linear_model.rds\n")
cat("- random_forest_model.rds\n")
cat("- income_distribution.png\n")
cat("- performance_by_department.png\n")
cat("- income_vs_experience.png\n")

How to work with R files

R provides comprehensive tools and environments for statistical computing:

R Development Environments

• RStudio - Most popular integrated development environment
• R GUI - Basic R interface included with installation
• Jupyter Notebooks - Interactive notebooks with R kernel
• Visual Studio Code - R extension for modern development
• Emacs ESS - Emacs Speaks Statistics package
• Vim-R - R support for Vim editor

Package Management

• CRAN - Comprehensive R Archive Network (main repository)
• Bioconductor - Bioinformatics packages
• GitHub - Development versions and specialized packages
• install.packages() - Install packages from CRAN
• devtools - Development tools and GitHub installation
• packrat/renv - Project-specific package management

Popular R Packages

• tidyverse - Collection of data science packages
• ggplot2 - Grammar of graphics for visualization
• dplyr - Data manipulation and transformation
• tidyr - Data tidying and reshaping
• readr/readxl - Data import and export
• stringr - String manipulation
• lubridate - Date and time handling
• caret - Classification and regression training

Statistical and Machine Learning Packages

• randomForest - Random forest algorithm
• e1071 - Support vector machines and other algorithms
• glmnet - Regularized linear models
• survival - Survival analysis
• nlme/lme4 - Mixed-effects models
• forecast - Time series forecasting
• cluster - Cluster analysis

Data Visualization

R excels in creating sophisticated visualizations:

• Base R graphics - Built-in plotting functions
• ggplot2 - Grammar of graphics approach
• plotly - Interactive web-based plots
• lattice - Trellis graphics for multivariate data
• shiny - Interactive web applications
• leaflet - Interactive maps
• DT - Interactive data tables

Statistical Analysis Capabilities

R provides comprehensive statistical functionality:

• Descriptive statistics - Summary statistics and distributions
• Hypothesis testing - t-tests, ANOVA, chi-square tests
• Regression analysis - Linear, logistic, polynomial regression
• Time series analysis - ARIMA, seasonal decomposition
• Survival analysis - Kaplan-Meier, Cox models
• Multivariate analysis - PCA, factor analysis, clustering
• Bayesian statistics - MCMC, Bayesian inference

R Markdown and Reproducible Research

R integrates with R Markdown for reproducible research:

• R Markdown - Combine code, results, and narrative
• knitr - Dynamic report generation
• Bookdown - Authoring books and long-form documents
• Blogdown - Creating websites and blogs
• Xaringan - HTML presentations
• Flexdashboard - Interactive dashboards

Data Science Workflow

R supports the complete data science pipeline:

• Data import - Read from various file formats and databases
• Data cleaning - Handle missing values, outliers, inconsistencies
• Exploratory data analysis - Understand data patterns and relationships
• Feature engineering - Create and transform variables
• Modeling - Build predictive and inferential models
• Validation - Cross-validation and model assessment
• Communication - Reports, dashboards, and presentations

Common Use Cases

R is widely used for:

• Academic research - Statistical analysis in social sciences, psychology, economics
• Biostatistics - Clinical trials, epidemiology, genetics
• Business analytics - Customer analytics, market research, A/B testing
• Finance - Risk modeling, algorithmic trading, econometrics
• Data science - Machine learning, predictive modeling
• Quality control - Statistical process control, six sigma
• Survey analysis - Survey design and analysis
• Environmental statistics - Ecological modeling, climate analysis
• Social media analytics - Text mining, sentiment analysis