Bitcoin Price Predictor (CLI)

This project is a Command Line Interface (CLI) application that retrieves historical Bitcoin price data, stores it in a database, trains a predictive model, and generates future price predictions based on user-defined parameters.

The application is designed as an end-to-end data pipeline, covering data acquisition, persistence, model training, and prediction, all operated through a simple and interactive terminal interface.

Features

-Retrieves historical OHLCV data from a local CSV file and the CoinGecko public API -Stores retrieved data in a database for reuse and reproducibility -Allows flexible date range selection for both data fetching and model training -Trains a regression-based model using historical data -Predicts future Bitcoin prices for a user-defined number of days -Provides clear previews of fetched and training datasets directly in the CLI -Fully interactive and easy to run from the terminal

Installation

1. Clone the repository:

   git clone https://github.com/andreasoledadguerra/predict_BTC.git
   cd predict_BTC

2. It is recommended touse a virtual environment.

   python -m venv venv
   source venv/bin/activate   # Linux / macOS
   # venv\Scripts\activate    # Windows

3. Install the required dependencies:

   pip install -r requirements.txt

Prerequisites

Before running the application make sure you have:

• Docker installed and running
• Docker Compose available
• Pyhon 3.x installed

Dependencies

The core Python dependencies include:

• pandas -data manipulation and analysis.
• numpy -numerical computations.
• requests -HTTP requests to external APIs(CoinGecko).
• psycopg2(or equivalent) - PostgreSQL database driver.
• scikit-learn -regression model training and prediction.
• SQLAlchemy -database abstraction and ORM

The exact database driver depends on the database configured in Docker.

Usage

1. Start the Docker container: The database used by the application runs inside a Docker container and must be startes first:

docker compose up -d

Make sure the container is running before proceeding.

2. Run the CLI Application:

   python3 predict_BTC/scripts/fetch_coingecko.py

Follow the on-screen instructions to:

• Select date ranges
• Train the model
• Generate price predictions

Project Structure

• src/ : Core application logic
• src/api: A client for fetching historical Bitcoin price data from the CoinGecko API.
• src/database/ : Database-related files and persisted data.
• src/ml/ : Data models and prediction logic.
• src/config: Settings manager for database and API configurations using environment variables.
• src/pipeline: Data pipeline class.
• src/utils: Utility for converting human-readable dates to Unix timestamps used by APIs.
• docker/ : Docker-related configuration.
• docker/env/ : Docker environment variables.
• src/main.py : CLI script to fetch data and run predictions.
• .env : Environment variables (not committed).
• docker-compose.yml : Docker services definition (database)
• tools/: Scripts for processing and loading canonical data into the PostgreSQL database.
• plots: Data visualization.
• plots/: Save generated plots.
• BTCUSD_1d_Binance.csv: historical OHLCV data.
• .gitignore : Git ignore rules.
• requirements.txt : Python dependencies.
• README.md : Project documentation.
• LICENSE : Project license.

Contributing

Contributions are welcome! Please fork the repository and submit a pull request with your changes.

License

This project is licensed under the terms specified in the LICENSE file.