Skip to contents

Getting Started with AgenticWaves

Source: vignettes/getting-started.Rmd
getting-started.Rmd

Getting Started with AgenticWaves

Welcome to AgenticWaves, a revolutionary R package that combines autonomous AI agents with advanced financial network analysis. This guide will help you get started with the core features and functionality.

Installation

From GitHub 

# Install from GitHub (recommended)
devtools::install_github("avishekb9/AgenticWaves")

# Or download and install locally
# git clone https://github.com/avishekb9/AgenticWaves.git
# cd AgenticWaves
# source("install.R")

Load the Package

When you load AgenticWaves, you’ll see a welcome message with quick start information.

Quick Start

Interactive Launcher

The easiest way to explore AgenticWaves is through the interactive launcher:

This opens an interactive menu with 11 different analysis options:

  1. 📱 Launch Interactive Dashboard - Full Shiny application
  2. 🤖 Quick Autonomous Analysis - AI-powered analysis with sample data
  3. 🎬 Run Complete Demo - Full demonstration with all features
  4. 🕸️ Network Analysis Only - Focus on spillover networks
  5. 👥 Agent Simulation Only - Agent-based market modeling
  6. 📁 Load Custom Dataset - Upload your own data
  7. 🎨 Generate Visualization Gallery - Publication-quality plots
  8. 📄 Create Sample Report - Comprehensive analysis report
  9. 🔍 System Diagnostics - Check system status
  10. 📚 View Documentation - Package help
  11. 🧪 Test All Functions - Comprehensive testing

Shiny Dashboard

For interactive analysis, launch the comprehensive Shiny dashboard:

The dashboard includes: - Data upload and processing - Autonomous AI analysis configuration - Interactive network visualization - Agent simulation setup - Real-time spillover analysis - Visualization gallery - Report generation

Basic Analysis Workflow

1. Load Sample Data

AgenticWaves includes built-in sample datasets for immediate testing:

# Load global stock market data
data <- get_sample_data("global_markets", n_assets = 10, n_periods = 500)

# Other options:
# data <- get_sample_data("crypto")          # Cryptocurrency data
# data <- get_sample_data("commodities")     # Commodity market data

2. Create an Autonomous AI Agent

Create an AI agent that can autonomously analyze your data:

# Create an explorer-type agent
agent <- create_autonomous_agent("explorer")

# Other agent types:
# agent <- create_autonomous_agent("optimizer")   # Optimization-focused
# agent <- create_autonomous_agent("predictor")   # Prediction-focused

3. Run Autonomous Analysis

Let the AI agent analyze your data autonomously:

results <- agent$analyze_autonomously(
  data = data,
  objective = "exploration",
  autonomy_level = 3
)

# View insights
print(results$insights)

# Check performance
print(results$performance_score)

4. Agent-Based Market Simulation

Create a population of trading agents and simulate market dynamics:

# Create diverse agent population
agents <- create_enhanced_agent_population(
  n_agents = 500,
  behavioral_heterogeneity = 0.7,
  wealth_distribution = "pareto"
)

# Run market simulation
sim_results <- simulate_enhanced_market_dynamics(
  agents = agents,
  asset_data = data,
  n_periods = 1000,
  network_effects = TRUE
)

# View simulation summary
print(paste("Final wealth Gini:", round(sim_results$final_wealth_gini, 3)))
print(paste("Average agent return:", round(mean(sim_results$agent_returns) * 100, 2), "%"))

5. Spillover Network Analysis

Analyze dynamic spillover effects and detect contagion episodes:

# Calculate dynamic spillovers
spillover_results <- calculate_dynamic_spillover_networks(
  simulation_results = sim_results,
  window_size = 100,
  significance_level = 0.05
)

# View spillover statistics
print(paste("Average spillover:", round(mean(spillover_results$total_spillover), 2), "%"))
print(paste("Peak spillover:", round(max(spillover_results$total_spillover), 2), "%"))
print(paste("Contagion episodes:", nrow(spillover_results$contagion_episodes)))

6. Advanced Contagion Detection

Use multiple methodologies for robust contagion detection:

# Detect contagion episodes using multiple methods
contagion_results <- detect_contagion_episodes(
  spillover_results = spillover_results,
  detection_methods = c("threshold", "regime", "correlation", "volatility")
)

# View episode characteristics
print(contagion_results$episode_characteristics)

7. Publication-Quality Visualizations

Generate professional visualizations for papers and presentations:

# Create comprehensive dashboard
plots <- generate_publication_dashboard(
  simulation_results = sim_results,
  spillover_results = spillover_results,
  output_dir = "publication_output",
  save_plots = TRUE
)

# Individual network plot
network_plot <- plot_enhanced_network(
  spillover_results$network_results,
  layout = "stress",
  node_size_var = "degree",
  color_scheme = "viridis"
)

Data Quality and Preprocessing

Validate Your Data

Before analysis, check data quality:

# Assess data quality
quality_report <- validate_data_quality(data)

print(paste("Quality score:", quality_report$quality_score, "/100"))
print(paste("Quality level:", quality_report$quality_level))

# View recommendations
if(length(quality_report$recommendations) > 0) {
  cat("Recommendations:\n")
  for(rec in quality_report$recommendations) {
    cat("•", rec, "\n")
  }
}

Process Financial Data

Clean and preprocess your data:

# Process data with outlier removal
processed_data <- process_financial_data(
  data = data,
  remove_outliers = TRUE,
  standardize = FALSE
)

# Detect asset classes automatically
asset_classes <- detect_asset_classes(processed_data)
print(asset_classes)

Custom Data Analysis

Load Your Own Data

AgenticWaves can analyze your custom datasets:

# Load CSV data
custom_data <- read.csv("your_financial_data.csv")

# Process for AgenticWaves
processed_custom <- process_financial_data(custom_data)

# Validate quality
quality <- validate_data_quality(processed_custom)

# Run analysis
custom_agent <- create_autonomous_agent("optimizer")
custom_results <- custom_agent$analyze_autonomously(processed_custom)

Next Steps

  • Explore the Interactive Dashboard for hands-on analysis
  • Read the Agent-Based Modeling guide for detailed ABM information
  • Check the Network Analysis article for advanced spillover techniques
  • Review the Visualization Guide for publication-ready plots
  • Try the Advanced Analysis tutorial for complex workflows

Getting Help

For a quick demonstration of capabilities, run:

# Generate visualization gallery
generate_visualization_gallery()

# Create sample report
create_sample_report()

# Run complete demo
run_complete_demo()

This will create comprehensive examples showing all major features of AgenticWaves.