FinNet - Financial Social Network Analysis

A comprehensive Financial Social Network Analysis platform with a modern CustomTkinter GUI.

✨ Features

• 📊 100% Free Data Sources - yFinance, Reddit, Google News, RSS feeds
• 🕸️ Complete SNA Coverage - All 6 units of Social Network Analysis curriculum
• 🧠 GCN from Scratch - Pure PyTorch implementation (no PyTorch Geometric)
• 🎨 Modern GUI - Beautiful dark theme with CustomTkinter
• 📈 Real-time Analysis - Live sentiment and network metrics

🚀 Installation

pip install finnet

📋 Quick Start

Launch GUI

finnet

CLI Analysis

finnet --cli --tickers AAPL,MSFT,GOOGL --period 6mo

Python API

from finnet.data_collection import StockDataCollector, RedditJSONScraper
from finnet.network_analysis import FinancialNetworkBuilder, CommunityDetector
from finnet.machine_learning import Node2VecEmbed, GCNLayer

# Collect data
collector = StockDataCollector()
stocks = collector.get_multiple_stocks(['AAPL', 'MSFT', 'GOOGL'])

# Build network
builder = FinancialNetworkBuilder()
G = builder.build_correlation_network(stocks, threshold=0.5)

# Detect communities
detector = CommunityDetector()
communities = detector.detect_louvain(G)

# Generate embeddings
n2v = Node2VecEmbed(dimensions=64)
n2v.fit(G)
embeddings, nodes = n2v.get_all_embeddings()

📚 Syllabus Coverage

🛠️ Project Structure

finnet/
├── data_collection/     # yFinance, Reddit, News, RSS
├── data_processing/     # Sentiment, cleaning, features
├── network_analysis/    # Units 1-5 algorithms
├── machine_learning/    # GCN, Node2Vec (from scratch)
├── visualization/       # Matplotlib, Plotly, PyVis
├── gui/                 # CustomTkinter 9-tab interface
│   └── tabs/           # Dashboard, Input, Network, etc.
└── reports/            # PDF, Excel export

📦 Dependencies

Core: yfinance, networkx, pandas, numpy, torch, customtkinter

Optional: gensim, transformers, pyvis, plotly, fpdf2

📄 License

MIT License - see LICENSE