NHANES Hypertension Prediction Project

Overview

This project analyzes NHANES (National Health and Nutrition Examination Survey) data to predict hypertension using various lifestyle factors. The pipeline consists of three main components:

1. Data Combiner: Combines and filters raw NHANES CSV files
2. Data Autofiller: Handles missing data and creates composite features
3. Statistics Creator: Generates comprehensive statistical analysis reports

Features

Data Processing Pipeline

• Combines multiple NHANES dataset files by year
• Filters relevant health and lifestyle variables
• Handles missing data through intelligent autofilling
• Creates composite health indicators
• Generates statistical reports and visualizations

Key Health Indicators

• Blood Pressure Measurements
• Cholesterol Levels (Total, HDL, LDL)
• Smoking History
• Alcohol Consumption
• Diet and Nutrition
• Mental Health Factors
• Weight History
• Cardiovascular Health

Project Structure

project/
├── data_combiner/
│   ├── core/
│   └── utils/
├── data_autofiller/
│   ├── core/
│   └── services/
├── statistics_creator/
│   ├── analyzers/
│   └── visualizers/
├── questions/
│   └── *.json
└── data/
    ├── raw/
    └── processed/

Installation

# Install data_combiner dependencies
pip install -r data_combiner/requirements.txt

# Install statistics_creator dependencies
pip install -r statistics_creator/requirements.txt

Usage

1. Data Combining

from data_combiner import DataCombiner

# Combine and filter NHANES data files
combiner = DataCombiner(input_files)
combiner.combine_data()

2. Data Autofilling

from data_autofiller.services import AutofillService

# Initialize service
service = AutofillService(
    data_reader=FileDataReader(),
    question_repository=FileQuestionRepository(),
    rule_engine=DefaultRuleEngine(),
    config=autofill_config
)

# Process files
service.process_files(input_files, output_dir)

3. Statistical Analysis

from statistics_creator import StatisticsCreator

# Generate statistical reports
statistics_creator = StatisticsCreator(data_loader, analyzers, visualizers)
results = statistics_creator.run_analysis(data_path)

Configuration

The project uses JSON configuration files in the questions/ directory to define:

• Required variables
• Valid value ranges
• Skip patterns
• Autofill rules
• Composite variable formulas

Example configuration: