Water Quality Testing Data Analysis

Statistical analysis and predictive modeling of water quality parameters using Python.

Overview

This project analyzes a dataset of 500 water samples across five quality parameters to explore relationships between water quality indicators and build predictive models for conductivity.

Key findings:

• Strong positive correlation (r = 0.705) between pH and dissolved oxygen levels
• Multi-parameter linear regression model predicts conductivity from pH, temperature, turbidity, and dissolved oxygen
• OLS regression confirms statistically significant relationships between several parameter pairs (p < 0.05)

Dataset

The dataset (data/water_quality_testing.csv) contains 500 samples with the following parameters:

Parameter	Unit	Range
pH	pH units	6.83 - 7.48
Temperature	°C	20.3 - 23.6
Turbidity	NTU	3.1 - 5.1
Dissolved Oxygen	mg/L	6.0 - 9.9
Conductivity	µS/cm	316 - 370

Project Structure

water-quality-testing-data-analysis/
├── data/
│   └── water_quality_testing.csv       # Water quality dataset (500 samples)
├── notebooks/
│   └── water_quality_analysis.ipynb    # Main analysis notebook
├── .gitignore
├── CODE_OF_CONDUCT.md
├── LICENSE
├── README.md
└── requirements.txt

Getting Started

Prerequisites

• Python 3.8+
• pip

Installation

git clone https://github.com/ChanMeng666/water-quality-testing-data-analysis.git
cd water-quality-testing-data-analysis
pip install -r requirements.txt

Usage

jupyter notebook notebooks/water_quality_analysis.ipynb

Run all cells (Kernel > Restart & Run All) to reproduce the full analysis.

Analysis Contents

The notebook covers the following topics:

1. Data Loading and Inspection - Load dataset, examine structure and summary statistics
2. Distribution Analysis - Histograms with KDE for all parameters
3. Correlation Analysis - Correlation matrix heatmap and pair plots
4. pH vs Dissolved Oxygen - Deep dive into the strongest correlation
5. Parameter Relationships - Regression plots for multiple parameter pairs
6. Predictive Modeling - Linear regression for conductivity prediction (two-feature and multi-parameter models)
7. Statistical Modeling (OLS) - Ordinary least squares regression with statsmodels for statistical inference
8. Conclusions - Summary of key findings

Built With

• pandas - Data manipulation and analysis
• NumPy - Numerical computing
• Matplotlib - Static plotting and visualization
• seaborn - Statistical data visualization
• scikit-learn - Linear regression modeling
• statsmodels - OLS regression and statistical testing

License

This project is licensed under the MIT License. See LICENSE for details.

Author

Chan Meng - GitHub · LinkedIn · Website