Functional Fitness Data Modeling using Python

This repository contains a Jupyter notebook focusing on modeling a synthetic dataset representing high-intensity fitness program performance using Python.

Project Overview

This project aims to simulate a dataset that mirrors real-world phenomena related to high-intensity workout programs. It explores four key variables:

1. Blood lactate concentration
2. Heart rate during the workout
3. Subjective difficulty level experienced during the workout
4. Caloric expenditure

File Structure

• project.ipynb: Jupyter notebook containing the detailed process of dataset simulation, variable investigation, analysis, interpretation, and references.
• readme.md: This file, providing an overview of the project, its objectives, and the tools used.
• .gitignore: Used to keep the repository tidy, and avoid accidental file upload.

Project Structure

The project.ipynb file within this repository is organized into the following sections, which are stylised using markdown for ease of usage:

• Imports
• Project Brief
• Introduction
• Dataset Overview
• Simulation Steps & Design
• Implementation of Simulation
• Analysis and Insights
• Conclusions
• References

Tools and Libraries

This project was developed using:

• Visual Studio Code
• Jupyter Notebooks (via Anaconda)
• Python 3.11.5
• Libraries used:
  • Pandas
  • Matplotlib
  • NumPy
  • Seaborn

Running the Jupyter Notebook

To run the Jupyter Notebook included in this repository, follow these steps:

Clone the Repository

git clone https://github.com/nexlanglxm/data-modelling.git

cd project.ipynb

Install Required Dependencies

Ensure you have Python installed, preferably Python 3.11, along with the necessary libraries. You can install the required libraries using pip:

pip install pandas matplotlib numpy seaborn

Open the Notebook

Open the Jupyter Notebook using Jupyter Notebook, Jupyter Lab, or Visual Studio Code.

Execute Cells

Inside the notebook, execute each cell sequentially by clicking on them and pressing Shift + Enter to run the code cells.

Data Generation Methodology

The synthetic dataset was generated following specific distributions for each variable.

• Blood Lactate Concentration (BLC):

  • Description: A continuous variable measuring lactate in the blood during high-intensity workouts.
  • Ranges: Resting blood lactate: Typically up to 2.0 mmol/L; during intense exercise, levels may elevate to 15-25 mmol/L.
  • Simulation: Synthetic data for resting and exercise lactate levels were generated using numpy's random number generation with constrained values to align with realistic physiological levels.

• Heart Rate:

  • Description: A continuous variable reflecting physiological response during workouts.
  • Ranges: Target heart rate zones vary from 70-85% of maximum heart rate for different intensities.
  • Simulation: Synthetic heart rate data was created using numpy's random normal distribution centered around an average of 142 beats per minute.

• Subjective Difficulty Level (SDL):

  • Description: An ordinal variable expressing perceived difficulty levels, which was adapted to be numerical using Borg's Perceived Rate of Exertion.
  • Ranges: Subjective difficulty levels were simulated using numpy's random normal distribution centered around a mean value of 142.
  • CR10 Scaling: Heart rate data was mapped to CR10 scaling categories to simulate subjective difficulty levels based on exercise intensity.

• Caloric Expenditure:

  • Description: A continuous variable measuring energy expended during workouts.
  • Ranges: Caloric expenditure was estimated using Metabolic Equivalent of Task (MET) values for circuit training and an average individual weight of 70 kilograms.
  • Simulation: Synthetic data for caloric expenditure was generated using numpy's random normal distribution with a mean value estimated from the MET values.

These methodologies were used to generate synthetic data for the variables, ensuring realistic ranges and distributions aligned with known physiological responses during high-intensity workouts.

Acknowledgements

References and sources used in this project are cited during and listed in their entireity at the end of the project.ipynb file.