Data Structures & Algorithms lab in Python with step-by-step visualization in Streamlit + Graphviz.

Problem

Learning data structures only through the console becomes abstract: it’s hard to see the real internal state (pointers, buckets, rotations, traversals). I wanted a way to practice interview-style but with immediate visual feedback.

What I built

An interactive lab that simulates operations step by step:

Executes operations line by line (with # comments)
Generates navigable snapshots (Prev/Next/Reset)
Renders the state with Graphviz and also shows the state as text
Includes linear, hash, and tree modules (BST/AVL/LLRB)

Screenshots

PyDSA Lab - Stack PyDSA Lab - Hash Set PyDSA Lab - AVL Tree

Architecture

pages/: UI por módulo (Streamlit)
core/structures/: data structure implementations (DS)
core/algos/: parseo + handlers + build_steps (simulación)
core/render/: *_to_dot (Graphviz)
core/stepper.py: snapshot navigation
tests/: smoke tests + ops/render tests

Challenges & tradeoffs

Unifying contracts: consistency between Step (values/buckets/etc.) and what tests/pages consume.
items() in some structures as a generator: great for performance, but in tests/UI it’s sometimes better to materialize with list(...).
Keeping algos decoupled from render by passing dot_builder from the page.

Results / Impact

A repeatable workflow to practice DS with real visualization.
Tests + hooks (ruff/pre-commit) keep quality high and prevent regressions during refactors.

What I’d improve next

Export snapshots to JSON to save/load scenarios
Autoplay mode (play/pause + speed)
More modules: Heap / Trie / DSU / LRU Cache
Better UI: operation history and per-structure filters