Algorithm Lab

A Deno + TypeScript playground for exploring algorithmic complexity, performance characteristics, and type-safe API design.

Getting started

deno task start

An interactive CLI lets you select which algorithms to benchmark, set iteration counts, and view results in a live-updating table.

Big-O complexity chart

Algorithms

Search

Algorithm	Complexity	Notes
Linear search	O(n)	Works on any array
Binary search	O(log n)	Requires SortedArray<T> (enforced by types)
Native search	O(n)	Delegates to Array.findIndex

Sort

Algorithm	Complexity	Stable	Mutates
Native sort	O(n log n)	Yes	No*
Merge sort	O(n log n)	Yes	No
Insertion sort	O(n²)	Yes	No
Selection sort	O(n²)	No	No

*Wraps Array.sort (TimSort in V8) but copies the input first to avoid mutation.

Benchmark setup

• Multiple dataset sizes per category (search: 10K–1M, sort: 1K–50K).
• JIT warmup phase before measurement.
• Per-iteration timing via performance.now() with min/avg/max reporting.
• Randomized input for sort, worst-case (last element) for search.

TypeScript design focus

This project uses TypeScript as a design and modeling tool:

• Branded types — SortedArray<T> and Index encode domain invariants at the type level, making illegal states unrepresentable.
• Generic algorithm interfaces — SearchAlgorithm<T, A> and SortAlgorithm<T> define contracts that all implementations conform to.
• Factory functions — each algorithm is a factory returning an interface implementation, keeping instantiation uniform.