Stock Analysis Django Project

This project is a Django-based web application for advanced stock analysis. It provides two different analysis methods:

1. Heuristic Analysis: Combines technical indicators (RSI, MACD, Bollinger Bands), fundamental data, and sentiment analysis (using nltk's VADER and additional headlines via NewsAPI) to compute support/resistance levels, suggested buy/sell prices, and an overall recommendation.

2. LSTM Analysis: Uses a deep learning LSTM model (built with TensorFlow/Keras) to forecast the next day’s closing price based on historical data and provides a BUY/HOLD recommendation based on a 1% threshold. This method uses 10 years of historical data and trains for 20 epochs.

Additionally, users can choose to run both methods at once and compare the results.

Features

• Technical Analysis:

  • Computes technical indicators (RSI, MACD, Bollinger Bands) from historical data.
  • Calculates support levels, resistance levels, highest hit price, and suggested buy/sell prices.

• Fundamental Analysis:

  • Retrieves fundamental data such as P/E ratio, EPS, and Market Cap (formatted in Indian crores).

• Sentiment Analysis:

  • Uses nltk's VADER to compute sentiment scores from news headlines fetched from both yfinance and NewsAPI.
  • Aggregates headlines for a robust sentiment analysis.

• LSTM Forecasting:

  • Implements an LSTM model to forecast the next day’s closing price.
  • Uses 10 years of historical data for training and predicts a BUY/HOLD signal based on a 1% threshold.

• Modern UI:

  • A Bootstrap-based frontend provides a user-friendly interface.
  • Users can select the analysis method (Heuristic, LSTM, or Both) and view trending symbols.

Installation

1. Clone the Repository:

   git clone https://github.com/yourusername/stock-analysis-django.git
   cd stock-analysis-django

Create and Activate a Virtual Environment:

python -m venv venv source venv/bin/activate # On Windows use: venv\Scripts\activate

pip install django yfinance tensorflow scikit-learn pandas numpy matplotlib ta nltk requests Set Up Environment Variables (Optional):

For sentiment analysis, the application uses NewsAPI. Obtain a free API key from NewsAPI.org and set it in your environment:

On Windows (Command Prompt):

set NEWS_API_KEY=your_news_api_key On Windows (PowerShell):

powershell

$env:NEWS_API_KEY="your_news_api_key" On macOS/Linux:

export NEWS_API_KEY=your_news_api_key Apply Migrations:
python manage.py migrate Usage Run the Django Development Server:

python manage.py runserver Access the Application:

Open your web browser and navigate to http://127.0.0.1:8000/. On the home page, you can:

Enter a ticker symbol (e.g., TATAPOWER.NS). Select the analysis method via radio buttons: Heuristic Analysis LSTM Analysis Both Methods View a list of trending symbols. View the Report:

After submission, the application processes the request using the chosen method(s) and renders a detailed report.