Book Guide

Use this guide to move between Machine Learning for Trading, Third Edition and ml4t-backtest without guessing which notebook or case-study file maps to which production API.

The book teaches the concepts step by step. ml4t-backtest packages those ideas into reusable execution, reporting, and portfolio-simulation workflows.

How To Use This Guide

• start in the book when you want intuition, derivations, and side-by-side experiments
• start in the library docs when you want stable APIs and configurable workflows
• use this page to jump from a chapter or case study to the corresponding guide page

The most common reader journey is:

1. read the notebook that explains the trading or execution idea
2. use this guide to find the matching library workflow
3. read the user-guide page for the reusable configuration pattern
4. use the API Reference for exact classes and signatures

Chapter Map

Chapter 16: Strategy Simulation

Book path	What the book teaches	Library workflow	Docs page
16_strategy_simulation/02_order_types.py	order semantics, stop/limit behavior, and timing choices	order types and trigger behavior	Order Types
16_strategy_simulation/03_execution_semantics.py	next-bar vs same-bar timing, stop handling, and execution assumptions	execution-mode, trigger, and stop configuration	Execution Semantics
16_strategy_simulation/04_configuration_profiles.py	framework-style presets and reproducible backtest behavior	named config profiles and BacktestConfig	Configuration, Profiles
16_strategy_simulation/05_performance_reporting.py	trade analytics, tearsheets, and gross-vs-net reporting	BacktestResult, trade/fill/equity export, reporting helpers	Results & Analysis
16_strategy_simulation/06_framework_parity.py	why configurable semantics matter for reproducibility	strict profiles and parity validation philosophy	Profiles
16_strategy_simulation/08_rebalancing.py	weight-based portfolio management	target weights and rebalancing executors	Rebalancing

What changes when you move to the library:

• the book shows concepts one execution assumption at a time
• the library exposes those assumptions as named configuration knobs
• results become exportable, replayable, and comparable across frameworks

Chapter 17: Portfolio Construction

Book path	What the book teaches	Library workflow	Docs page
17_portfolio_construction/01_portfolio_metrics.py	allocator comparison and portfolio-level behavior	portfolio accounting and result metrics	Results & Analysis, Account Policies
17_portfolio_construction/08_library_comparison.py	comparing portfolio construction workflows	feed preparation, rebalancing, and configuration	Data Feed, Rebalancing
17_portfolio_construction/09_allocation_horse_race.py	realistic allocator comparison under costs	cost-aware execution and turnover analysis	Results & Analysis, Market Impact

What changes when you move to the library:

• allocators become reusable strategies or weight targets
• fills, turnover, and portfolio state become first-class reporting surfaces
• account policy becomes explicit instead of hidden in notebook assumptions

Chapter 18: Costs

Book path	What the book teaches	Library workflow	Docs page
18_costs/01_transaction_costs.py	commissions, slippage, and cost decomposition	commission and slippage configuration	Configuration, Market Impact
18_costs/02_market_impact.py	realistic impact assumptions and sensitivity to size	market-impact models and execution drag	Market Impact
18_costs/03_quote_aware_execution.py	bid/ask-aware execution and quote-side marking	quote-aware fills, bid/ask mark prices, quote-side execution	Execution Semantics, Data Feed

What changes when you move to the library:

• manual cost adjustments become configuration-backed execution models
• quote-aware assumptions become explicit and reproducible
• fills and trades carry enough context to audit cost drag directly

Chapter 19: Risk Management

Book path	What the book teaches	Library workflow	Docs page
19_risk_management/02_exit_strategies.py	stop-loss, take-profit, and trailing-stop design	position-rule composition	Risk Management
19_risk_management/03_position_sizing.py	position sizing and constraints	account policy, sizing, and exposure limits	Account Policies, Risk Management
19_risk_management/06_portfolio_limits.py	portfolio-level guardrails	portfolio limits and manager coordination	Risk Management

What changes when you move to the library:

• notebook exits become reusable rule objects
• portfolio-wide guardrails become configuration, not ad hoc if-statements
• risk and execution interact through explicit event ordering

Case-Study Pipeline Map

Most case studies use ml4t-backtest in the final simulation and reporting stages.

Case-study step	Typical file	Library workflow	Docs page
Backtest setup	case_studies/<study>/code/14_backtest.py	BacktestConfig, DataFeed, Engine, strategy bridge	Quickstart, Configuration, Data Feed
Performance reporting	case_studies/<study>/code/15_evaluation.py or equivalent	BacktestResult, trade/fill/equity export, downstream diagnostics	Results & Analysis
Portfolio/risk comparison	allocator or risk notebooks in Ch17-Ch19	account policy, turnover, limits, and rebalancing behavior	Account Policies, Rebalancing, Risk Management

Particularly strong bridges in the current materials:

• Ch16 case studies use setup.yaml and get_backtest_config() to turn notebook assumptions into reusable configuration
• NASDAQ-100 microstructure work demonstrates quote-aware execution and richer reporting
• portfolio-construction studies rely on rebalancing and turnover-aware result surfaces
• live-trading materials depend on keeping the same strategy shape before migrating to ml4t-live

From Notebook Code to Library API

Use this translation when moving from the book to reusable code:

Book pattern	Library equivalent
inline execution assumptions	BacktestConfig(...) or named profile
notebook-specific fill logic	ExecutionMode, fill-ordering, stop-fill configuration
manual portfolio loops	rebalancers and weight-target executors
ad hoc reporting frames	BacktestResult exports for trades, fills, equity, and portfolio state
cost tweaks in separate analysis cells	slippage, commission, and market-impact models in config
one-off state carried across notebook cells	stateful strategy patterns in reusable Strategy subclasses

What To Prioritize

These are the workflows most readers should prioritize:

• execution semantics and configuration profiles
• data-feed preparation and signal alignment
• result export and performance reporting
• risk-management rules and rebalancing
• quote-aware execution and market-impact modeling when realism matters

Related Docs

• Quickstart
• How It Works
• Execution Semantics
• Results & Analysis
• API Reference