AdvVisR - Advanced Data Visualisation in R

Project Overview

This repository hosts the AdvVisR project, which focuses on advanced data visualization techniques using R. The project analyzes a complex, simulated dataset regarding fitness activities, nutritional habits, and physiological metrics.

The core of the project involves a dual-stack approach:

1. Data Engineering (Python): A sophisticated simulation engine generates realistic synthetic data (Final_data_model.csv), enforcing logical correlations (e.g., Age vs. Metabolism, Diet Type vs. Macro Ratios).
2. Visualization (R): A suite of advanced, static, and interactive charts creates a narrative around the "fuel" (nutrition), the "machine" (workout mechanics), and the "results" (body composition).

Authors

• Cezary Kuźmowicz
• Wojciech Hrycenko

---

Repository Contents

1. Visualization & Analysis

File: adv_vis_r_kuzmowicz_hrycenko.r

Objective To visualize the fitness journey through three thematic lenses using advanced ggplot2 extensions:

• Part 1: The Fuel (Nutrition):
  • Treemap (treemapify): Analyzes Caloric Volume vs. Sugar Content across different diets (Keto, Paleo, Vegan).
  • Small Multiples Radar Chart: Compares nutritional balance (Proteins, Fats, Carbs, Sodium, Water) for each diet type.
• Part 2: The Machine (Process):
  • Alluvial Plot (ggalluvial): Traces the user flow from Experience Level → Workout Choice → Caloric Burn Efficiency.
  • Interactive Scatter (ggiraph): Explore the relationship between Heart Rate, Calories, and Duration with hover-over exercise details.
  • Circular Barplot: An "Anatomical Tachometer" visualizing targeted muscle groups.
• Part 3: The Results (Physiology):
  • Dumbbell Plot: Contrasts Resting Heart Rate vs. Max Heart Rate vs. Training Intensity.
  • Ridgeline Plot (ggridges): Visualizes "Mountains of Fat" – body fat distribution by gender and experience level.

2. Data Simulation Engine

File: Data/data_modelling_macros.py

Description A Python script designed to generate high-quality synthetic data for visualization purposes. It goes beyond simple random generation by applying "physics enforcement" logic:

• Population Balancing: Ensures balanced representation of difficulty levels.
• Macro Restructuring: Adjusts nutrient profiles (Proteins/Fats/Carbs) dynamically based on Diet Type (e.g., Keto has high fats/low carbs).
• Correlation Engineering: Enforces realistic biological relationships (e.g., BMI > 25 increases Heart Rate effort; Age reduces Max Heart Rate).

---

Course Requirements (Original)

Project Each student is obliged to submit group project (at most 2 students per group) regarding data visualisation. Topic and database can be selected freely as long as it utilizes techniques used for visualization in R (not necessarily discussed during classes). The basic requirements and conditions:

• Submitting project topic, dataset and group: End of the November,
• Performing presentation: last two classes in January,
• Submitting scripts and short description of the analysis.

The points (70%) can be gain from:

• Attractiveness of the chosen topic, quality of the insights infered from data,
• Advancement and innovativeness of R codes, effort invested in the project,
• Accuracy and visual aspect of the presentation,
• Attractiveness of the presentation.

Charts Documentation: Google Docs Link

---

Technologies and Libraries

The project utilizes a mix of R for visualization and Python for data preparation:

R (Visualization)

• tidyverse (ggplot2, dplyr): Core data manipulation and plotting.
• ggalluvial: For flow diagrams (Alluvial plots).
• treemapify: For creating treemaps.
• ggiraph: For adding interactivity to ggplot2 geometries.
• ggridges: For ridgeline density plots.
• ggcorrplot: For visualization of correlation matrices.

Python (Data Gen)

• Pandas & NumPy: For data frame manipulation and numerical simulation.

Usage Instructions

1. Clone this repository.
2. Data Generation (Optional): Run Data/data_modelling_macros.py if you wish to regenerate the dataset with new random seeds.
3. Visualization: Open adv_vis_r_kuzmowicz_hrycenko.r in RStudio.
4. Ensure the working directory is set correctly (setwd(...)).
5. Run the script to generate the plots in the RStudio Viewer.