🧑‍🔬 Capstone Mini Project - NHANES 2020 Body Measurement Analysis

A data-driven exploration of adult body metrics from the National Health and Nutrition Examination Survey (NHANES) 2020.

---

📊 Project Overview

This project analyses body measurements of US adults from the NHANES 2020 survey, with a particular focus on comparing males and females across several body composition metrics:

• BMI (Body Mass Index)
• Waist-to-Height Ratio (WHtR)
• Waist-to-Hip Ratio (WHR)

Utilizes powerful Python tools: NumPy, SciPy, Matplotlib, Seaborn, and Pandas.

---

🗂️ Dataset

• Source: gagolews/teaching-data (NHANES Adult BMX 2020)
• Files:
  • nhanes_adult_male_bmx_2020.csv
  • nhanes_adult_female_bmx_2020.csv
• Columns:
  Weight | Height | Upper Arm/Leg Length | Arm/Hip/Waist Circumference

ℹ️ Data loaded directly from GitHub with np.genfromtxt. Rows with missing values are dropped before analysis.

---

🧩 Project Structure

Capstone_mini_project.ipynb   # Main Jupyter analysis
README.md                     # This file

---

🧠 Analysis Workflow

Section	Description
1	Importing libraries and loading data
2	Histogram comparison: male vs. female weight
3	Box-and-whisker: weights by sex
4	Descriptive statistics & distribution shape
5	Computing BMI
6	Z-score standardization (female dataset)
7	Pairplot matrix + correlation analysis (Pearson/Spearman)
8	Calculating WHtR and WHR
9	Boxplots: WHtR & WHR by sex
10	Pros & cons of BMI, WHtR, WHR
11	Extreme BMI participants (lowest/highest 5)

---

✨ Key Findings (click to expand)

• Males were heavier on average
  Greater variability in weight compared to females.
• Both distributions are right-skewed:
  Most participants fall in lower weight ranges, with a tail of high values.
• Weight and BMI are strongly correlated:
  • Correlation coefficient ~0.9+ (Pearson).
• Males had substantially higher WHR:
  • Consistent with abdominal fat distribution patterns.
• Height weakly correlated with waist/hip circumference:
  • Height alone is a poor predictor of body composition here.
• Highest BMI individuals:
  • Above average on all circumference measures; average or below for height.

---

⚙️ Tech Stack

• Python 3.7+
• NumPy — numerical computing
• Matplotlib — data visualization
• SciPy — stats/correlation/skewness/kurtosis
• Seaborn — advanced pairplots
• Pandas — used for DataFrame creation

---

How to start

1. Clone this repo:

   git clone https://github.com/tomato9553-bit/Capstone-project.git
   cd Capstone-project

2. Install dependencies:

   pip install numpy matplotlib scipy seaborn pandas

3. Open the Jupyter notebook:

   jupyter notebook Capstone_mini_project.ipynb

💡 Requires an active internet connection to fetch the dataset.

---

⚠️ Limitations

• Missing rows are dropped (not imputed)
• Dataset limited to US adults (2020) — results may not generalize
• Measurement variability (waist/hip ratios) may affect reliability

---

⏳ Timeline

• Initiated: 08-04-2026
• Completed: 15-04-2026

---

👤 Author

M. Gopal
Data Science Enthusiast

---

✅ Project Checklist

• Data import/validation
• Exploratory analysis
• Metrics calculation (BMI, WHtR, WHR)
• Visualization & stats
• Summarize findings
• Write-up complete

---

Have questions or suggestions? Open an issue or pull request!