design-patterns

Design Patterns

Overview

Design patterns are time-tested solutions to common problems in software design. They encapsulate best practices for creating flexible, scalable, and maintainable systems. While design patterns don't add new functionality, they standardize how we solve design challenges — making systems easier to read, extend, and debug.

In algorithmic trading, design patterns are often used to structure trading engines, risk modules, strategy plug-ins, and message-based communication.

---

Status: 🟢 Mandatory

Who should learn this?
✅ Every software engineer who works with object-oriented or modular systems
✅ Especially important in backend development, distributed systems, and plugin-based architectures
✅ Strongly advised for anyone contributing to collaborative codebases or frameworks

---

Learning Objectives

• Recognize and apply common object-oriented design patterns
• Identify code smells and replace ad hoc logic with standard solutions
• Distinguish between creational, structural, and behavioral patterns
• Understand SOLID principles and their connection to pattern design
• Use language-specific implementations of patterns (e.g., Python, C++)

---

Key Concepts

• Creational Patterns – Control object creation
  • Singleton, Factory, Abstract Factory, Builder, Prototype
• Structural Patterns – Define object composition
  • Adapter, Decorator, Composite, Proxy, Facade, Bridge, Flyweight
• Behavioral Patterns – Handle communication and control flow
  • Observer, Strategy, Command, Chain of Responsibility, State, Template, Visitor
• Anti-Patterns – Recognize poor design habits (e.g., God Object, Spaghetti Code)
• Pattern Selection – When and where to apply each type of pattern
• SOLID Principles – Foundational design principles behind many patterns

---

Study Materials

Books

Title	Author(s)	Description
Head First Design Patterns	Eric Freeman & Elisabeth Robson	Friendly, visual, and approachable introduction to classic OO patterns
Design Patterns: Elements of Reusable Object-Oriented Software	Gamma, Helm, Johnson, Vlissides (Gang of Four)	The original GoF book — foundational, but more theoretical
Refactoring to Patterns	Joshua Kerievsky	Shows how to evolve code into good design through incremental improvements
Clean Architecture	Robert C. Martin	Explores architectural boundaries, layers, and the role of patterns
Patterns of Enterprise Application Architecture	Martin Fowler	More complex patterns for distributed and enterprise-grade systems

---

Online Courses / Resources

Course or Series	Platform	Description
Design Patterns in Java	Coursera (by UBC)	Formal intro using Java with clear examples and UML
Design Patterns – Udemy	Udemy	Multiple hands-on pattern courses (language-specific and theory-focused)
Refactoring to Design Patterns	Pluralsight	Practical application of patterns using real-world refactoring examples
Head First Design Patterns Companion Site	O'Reilly	Visual recaps and examples for every pattern
Design Patterns in Python	GitHub / GitBook	Refactoring Guru — excellent visual guides in multiple languages
Design Patterns Explained Simply	GitHub	Repos with clean implementations of each pattern in C++ and Python
Software Engineering Daily – Design Pattern Episodes	Podcast	Interviews and breakdowns on real-world use of design patterns

---

Applications in Algorithmic Trading

Pattern	Use Case
Strategy Pattern	Encapsulate multiple trading strategies that can be swapped at runtime
Observer Pattern	Listen for market events, position changes, or risk alerts
Factory Pattern	Create new strategy/risk modules based on config files
Command Pattern	Represent trading actions as objects (e.g., order send, cancel, modify)
Decorator Pattern	Add logging, risk checks, or latency monitors without modifying core code
Singleton Pattern	Global access to config, logging, or market state objects
State Pattern	Represent trading state machines (e.g., warmup, active, cooldown)

---

Hands-On Projects

• Refactor a backtest engine to use the Strategy Pattern for plug-and-play logic
• Implement an Observer-based market data listener
• Design a simple Factory to load trading strategies from YAML
• Decorate a signal generator with logging, throttling, or risk constraints
• Use State or Command to simplify logic flow in multi-leg orders

---

Assessment

• Identify at least 10 patterns in your own or public codebases
• Implement 5 core patterns in your preferred language
• Refactor an existing codebase to use Strategy, Factory, or Observer
• Create a UML diagram from a pattern-heavy module

---

Next Steps

• Refactoring
• Software Architecture
• Clean Code