By Jamal Noah Chester-Morris Jan 2024
This project is a Probability Distribution application that allows users to input stock ticker symbols and select a time frame. The application then renders a graph of the historical percentage returns based on the selected company and time frame. The time frames include 1D for daily % return, 1W for weekly % return, and 1M for monthly % returns. All probability distributions are calculated using data from the past 10 years. This is managed in /api/views.py, where the application connects to the unofficial Yahoo Finance API using a Python library.
The data is processed to include percentage rate of change values. These values populate a histogram, and the application calculates a probability for the returns to be within a specified range. The probability is calculated by determining the count (the number of times an event occurs) - in this case, the number of times a stock price return falls within that range. The probability is then derived as follows: probability = number of times a specific event occurs / total number of events.
- Backend: Django
- Frontend: React & Tailwind
Using Python 3.12.1, Node v20.11.0.
To get started with the application:
- Navigate to the application directory:
cd application/ - Create a virtual environment:
py -m venv venv - Install requirements:
pip install -r requirements.txt - Run the server:
py manage.py runserver
The application should not require any extra setup as the JavaScript and CSS build files have already been compiled. I have opted for Vite.js to manage my build pipeline. The frontend/ directory houses a React application built with Vite.js. If you try to use npm run dev and connect to the Django DRF API endpoint, you may encounter CORS issues. To avoid this, you can modify vite.config.js. However, as this is a small project I coded in a day, I decided to build the frontend directly into my Django project and serve the compiled .jsx files from my static folder.
NOTE: NOT SUITABLE FOR PRODUCTION