GlobalEconoSim - Global Trade Network Simulation

This project simulates international trade networks and their evolution over time, accounting for geographic distance, political relationships (friendship), tariffs, and transaction costs.

Deployed App: https://tradesimulation.streamlit.app/

Key Features

• Geographic Distance: Countries located closer to each other have lower transaction costs
• Friendship Dynamics: Political relationships affect tariffs and evolve as trade happens
• Network Statistics: Comprehensive analysis of network structure including:
  • Clustering coefficients
  • Centrality measures (degree, betweenness)
  • Network diameter and path lengths
  • Community detection

Interactive Simulation

The project includes a full-featured interactive Streamlit application that allows you to:

• Configure simulation parameters (countries, political blocs, tariffs, etc.)
• Run simulations and visualize trade networks in real-time
• Apply policy shocks and observe how they propagate through the network
• Compare network simulation with Mean-Field Approximation (MFA)
• Analyze economic metrics like GDP, poverty rates, and inequality

Running the Interactive App

1. Install the required packages:

   pip install streamlit networkx numpy pandas plotly~=5.19 scipy

2. Run the Streamlit app:

   streamlit run trade_app.py

3. The app will open in your web browser at http://localhost:8501

Simulation Parameters

• Countries: Control the number of countries in the simulation (4-80)
• Political blocs: Group countries into alliance blocs with preferential trading
• Tariff gap: Set how much higher inter-bloc tariffs are compared to intra-bloc tariffs
• Two-good world: Enable comparative advantage with countries producing two goods with different efficiency
• Policy shock: Apply tariff changes to specific countries and observe ripple effects

Components

• Trade Network: Directed graph with countries as nodes and trade relationships as edges
• Friendship Matrix: Tracks political relationships between countries
• Transaction Costs: Partially based on geographic distance between countries
• Tariffs: Based on political relationships (friendship matrix)

Simulation Metrics

The simulation tracks and visualizes:

1. Economic Indicators:

   • Average tariffs
   • Transaction costs
   • Friendship levels
   • GDP growth
   • Poverty rates
   • Inequality (Gini coefficient)

2. Network Structure:

   • Number of trade relationships
   • Network diameter
   • Clustering coefficient
   • Community formation

3. Trade Hubs:

   • Countries with high betweenness centrality
   • Major players in the global trade network

File Structure

• trade_app.py - Main Streamlit application
• model.py - Core simulation primitives and engine
• visualization.py - Plotting and visualization functions
• analytics.py - Statistical analysis and advanced metrics
• trade_stats.py - Functions for analyzing trade network statistics

Requirements

• Python 3.x
• NetworkX
• NumPy
• Matplotlib
• Pandas
• Plotly
• Streamlit
• Scipy