R Statistical Data Analysis Project

This repository contains a comprehensive collection of statistical data analysis laboratories using R. Each lab focuses on different aspects of data exploration, visualization, analysis techniques, and machine learning models.

Project Structure

.
├── .gitignore                 # Git ignore configuration
├── LICENSE                    # MIT License
├── Lab1/                      # Laboratory 1 - Data Preprocessing & Exploratory Analysis
│   ├── Excel/                 # Directory for Excel files
│   ├── Univariate-Bivariate-Data/
│   │   └── main.R             # R script for univariate/bivariate analysis
│   └── main.R                 # Main script for Wesbrook dataset analysis
├── Lab2/                      # Laboratory 2 - Machine Learning Models
│   └── main.R                 # R script for predictive modeling
├── Lab3/                      # Laboratory 3 - Advanced Analytics
│   ├── main.R                 # R script for correlation analysis with Eggs dataset
│   ├── main_1.R               # R script for market basket analysis (association rules)
│   └── main_2.R               # R script for customer segmentation (clustering)
├── Lab4/                      # Laboratory 4 - Visualization & Statistical Plots
│   └── main.R                 # R script for advanced data visualization
└── data/                      # Data directory
    └── LifeCycleSavings_with_additional_columns.csv  # Extended dataset

Lab Descriptions

Lab 1: Data Preprocessing & Exploratory Analysis

This lab focuses on data preprocessing and exploratory analysis using two datasets:

1. LifeCycleSavings dataset:

• Pair plots to show dependencies between variables
• Histograms for univariate analysis
• Bar plots comparing countries on various metrics
• Age distribution analysis
• Scatter plots to explore relationships between variables

2. Wesbrook dataset:

• Data preprocessing and cleaning
• Missing values analysis and imputation
• Exploratory data visualization with boxplots and histograms
• Mosaic plots for categorical variables
• Feature scaling and normalization
• Stratified sampling for train/validation split

Lab 2: Machine Learning Models

This lab implements and evaluates various machine learning models for predicting donor behavior using the Wesbrook dataset:

• K-Nearest Neighbors (KNN)
• Naive Bayes
• Decision Trees
• Random Forests
• XGBoost
• Support Vector Machines (SVM)

Each model is evaluated using:

• Multiple cross-validation techniques (k-fold, LGOCV)
• Confusion matrices
• ROC curves and AUC metrics
• Parameter optimization

Lab 3: Advanced Analytics

This lab covers three distinct analytical approaches:

1. Correlation Analysis (Eggs dataset):

• Correlation between sales and price variables
• Matrix scatter plots for relationship visualization
• Time-series analysis of egg sales and prices
• Factor analysis affecting egg sales

2. Market Basket Analysis (Groceries dataset):

• Association rule mining with the Apriori algorithm
• Frequency analysis of purchased items
• Support, confidence, and lift metrics

3. Customer Segmentation (Mall Customers dataset):

• K-means clustering for customer segmentation
• Optimal cluster determination using elbow, silhouette, and gap statistic methods
• Demographic analysis of clusters
• Spending patterns and income distribution analysis

Lab 4: Statistical Visualization

This lab demonstrates various statistical visualization techniques in R:

• Histogram comparison with different settings
• Density plots and probability distributions
• Strip charts for group comparisons
• Box plots for distribution comparison
• Dual axis plots
• Pie charts for categorical data

Key Datasets

LifeCycleSavings Dataset

Contains savings and related economic data for various countries, including:

• Savings rates (sr)
• Population demographics (pop15, pop75)
• Average income (dpi)
• Income growth (ddpi)

Wesbrook Dataset

Contains donor information for fundraising analysis:

• WESBROOK indicator (donations >$1000 in recent fiscal year)
• TOTLGIVE (total donations)
• Demographic variables (MARITAL, SEX, etc.)
• Geographic and economic indicators (AVE_INC, DWEL_VAL, etc.)

Eggs Dataset

Analyzes egg sales data with related variables:

• Cases (egg sales)
• Various price indicators (Egg.Pr, Beef.Pr, Chicken.Pr, etc.)
• Seasonal factors (Month, Easter)

Groceries Dataset

Contains transaction data for market basket analysis:

• Multiple items purchased in each transaction
• Frequencies of item purchases

Mall Customers Dataset

Contains customer segmentation data:

• Customer demographics (Gender, Age)
• Income
• Spending Score

Getting Started

Prerequisites

• R (recommended version 4.0.0 or higher)
• RStudio (recommended for easier workflow)

Required R Packages

# Data manipulation
install.packages(c("tidyverse", "dplyr"))

# Visualization
install.packages(c("corrplot", "car", "plotly", "ggplot2", "rpart.plot", "factoextra"))

# Machine Learning & Statistics
install.packages(c("caret", "e1071", "ROCR", "rpart", "randomForest", "xgboost", "class"))

# Special Analytics
install.packages(c("arules", "stats"))

Running the Analysis

1. Clone this repository
2. Open the desired lab's main.R file in RStudio
3. Install any required packages
4. Execute the script to see the analysis results

Output

The scripts generate various visualization files in PNG format that are saved to the respective data directories. The analysis results include:

• Visualization plots (histograms, scatter plots, box plots, etc.)
• Model evaluation metrics (confusion matrices, ROC curves)
• Preprocessed datasets (saved as CSV files)
• Cluster analysis visualizations
• Association rules and patterns

Notes

• All R scripts use UTF-8 encoding
• For Lab1, the charts are automatically saved in the 'data/lab-1' folder
• Interactive plots in Lab3 require the plotly package
• Some models in Lab2 may take longer to run due to cross-validation

License

This project is licensed under the MIT License - see the LICENSE file for details.