Stock Price Prediction System

A comprehensive full-stack web application for real-time stock price analysis and prediction using 7 machine learning algorithms, supporting both US and Indian markets with live data fetching, historical analysis, and interactive visualization.

Key Features

• 7 ML Algorithms: Linear Regression, Decision Tree, Random Forest, SVM (basic models) + KNN, ARIMA, Autoencoder (advanced models)
• Real-time Data: US stocks via Finnhub API, Indian stocks via Upstox API with permanent storage fallback
• Modern Dashboard: React 18 + TypeScript + Tailwind CSS with interactive Recharts for 5-year historical analysis
• 1000+ Stocks: 1,001 total stocks available (501 US + 500 Indian), 936 used for training after filtering insufficient data
• Currency Support: Real-time USD/INR conversion via forex-python
• Smart Training: Percentage-based predictions with proper price conversion and confidence scoring
• Standalone Trainers: Independent training scripts for each model with progress tracking

Currency Conversion

Automatic USD/INR conversion for all stocks:

• Sources: Live forex APIs → Cached rates → Static rate (83.5 USD/INR)
• Display: Original currency + converted price shown for all stocks

Technology Stack

Backend: Flask 2.3.3, Python 3.8+, TensorFlow 2.20, scikit-learn 1.5.2, statsmodels 0.14.4, pandas, numpy
Frontend: React 18.3.1, TypeScript, Vite 6.4.0, Tailwind CSS, Radix UI, Recharts 2.15.2
APIs: Finnhub (US stocks), Upstox (Indian stocks), yfinance (historical data)

Offline Mode First

The system works completely offline without any API keys! This is the recommended way to get started.

What Works Offline

• Stock Information: 500 Indian + 501 US stocks (1,001 total) from permanent directory
• Historical Charts: Complete 5-year OHLCV data (2020-2024)
• ML Predictions: All trained models work with offline data (trained on 936 stocks with sufficient data)
• Search: Full-text search across 1,001 stocks
• Technical Indicators: 38 indicators calculated from historical data

What doesn't Work Offline

• Live Prices: Requires API keys (Finnhub for US, Upstox for India)

Installation Guide

Prerequisites

• Python: 3.8 or higher
• Node.js: 16 or higher
• Git: For cloning repository
• RAM: 8GB minimum, 16GB recommended
• Storage: 15GB free space

Windows Installation

1. Install Python

# Using winget (Windows Package Manager)
winget install Python.Python.3.11

# Or download from python.org
# Verify installation
python --version

2. Install Node.js

# Using winget
winget install OpenJS.NodeJS

# Verify installation
node --version
npm --version

3. Clone and Setup

# Clone repository
git clone https://github.com/ankit02327/stock-price.git
cd ml

# Backend setup
cd backend
python -m venv venv
venv\Scripts\activate
pip install -r requirements.txt

# Create empty .env file (REQUIRED)
type nul > .env

# Frontend setup (new terminal)
cd ../frontend
npm install

macOS Installation

1. Install Homebrew (if not installed)

/bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh)"

2. Install Python and Node.js

# Install Python and Node.js
brew install python@3.11 node

# Verify installations
python3 --version
node --version
npm --version

3. Clone and Setup

# Clone repository
git clone https://github.com/ankit02327/stock-price.git
cd ml

# Backend setup
cd backend
python3 -m venv venv
source venv/bin/activate
pip install -r requirements.txt

# Create empty .env file (REQUIRED)
touch .env

# Frontend setup (new terminal)
cd ../frontend
npm install

Linux Installation (Ubuntu/Debian)

1. Update Package List

sudo apt update

2. Install Python and Node.js

# Install Python and pip
sudo apt install python3.11 python3.11-pip python3.11-venv

# Install Node.js
curl -fsSL https://deb.nodesource.com/setup_18.x | sudo -E bash -
sudo apt-get install -y nodejs

# Verify installations
python3 --version
node --version
npm --version

3. Clone and Setup

# Clone repository
git clone https://github.com/ankit02327/stock-price.git
cd ml

# Backend setup
cd backend
python3 -m venv venv
source venv/bin/activate
pip install -r requirements.txt

# Create empty .env file (REQUIRED)
touch .env

# Frontend setup (new terminal)
cd ../frontend
npm install

Quick Start

1. Start Backend

cd backend
venv\Scripts\activate  # Windows
source venv/bin/activate  # macOS/Linux
python main.py

2. Start Frontend (New Terminal)

cd frontend
npm run dev

Access: http://localhost:5173 (Backend: http://localhost:5000)

Data Structure

Indian Stocks (500): Use ISIN codes (INExxxxxxxx) at permanent/ind_stocks/
US Stocks (501): Use ticker symbols (AAPL, MSFT, etc.) at permanent/us_stocks/

# Train: python backend/training/basic_models/{model}/trainer.py
# Status: python status.py
# Details: See [Training Guide](documentation/TRAINING.md)

Optional: Live Data Setup

Add API keys to backend/.env for real-time data (Finnhub for US, Upstox for India).

Guides: Offline Mode | Training Guide | Backend API | Upstox Integration

System Status

Check: python status.py | Test API: curl http://localhost:5000/health

Working: Data fetching, 5-year historical charts, search (1000+ stocks), currency conversion, interactive dashboard, 38 indicators, ML predictions (4/7 models trained)

Troubleshooting

Issue	Solution
Backend won't start	Verify permanent/ directory exists and create empty .env file
Frontend build errors	Run rm -rf node_modules package-lock.json && npm install
Python import errors	Activate venv: venv\Scripts\activate (Windows) or source venv/bin/activate (macOS/Linux)
Out of memory	Close other apps, ensure 8GB+ RAM, train one model at a time

Important Notes

• Educational Use: This application is for learning and research purposes only
• Investment Disclaimer: Stock predictions are inherently uncertain and should not be used as sole investment advice
• API Limits: Finnhub (60 calls/min free), Upstox tokens expire daily
• Offline First: System designed to work without internet or API keys

Quick Commands

python status.py  # Check status
python main.py    # Start backend
npm run dev       # Start frontend