Graph Algorithm Engine

Navigate Every
Route, Optimally.

A comprehensive flight network optimizer powered by Dijkstra, A*, Bellman-Ford, Floyd-Warshall, Yen's K-Paths, and more — running on a 300+ route global dataset.

Launch Optimizer
300+ Flight Routes
64 Airports
10 Algorithms
6 Continents
Live Route Network
Airport Node
Route
Highlighted Path
Scroll to zoom · Hover vertices and edges for details
Available Algorithms

Ten algorithms.
Every routing need covered.

Cheapest Route
Dijkstra on cost weights with automatic discounts applied. Finds the minimum-cost path using a priority queue and predecessor map.
O((V+E) log V) Greedy Single-source
Fastest Route
Dijkstra on duration weights. Minimises total flight time. Try JFK → SYD — the 19 h direct is fastest; the cheaper route via SIN takes 25 h.
O((V+E) log V) Greedy Single-source
A* Geo-Route
A* with Haversine heuristic. Try JFK → PER — A* routes via DXB (geographically east); Cheapest routes via SIN. Different paths, same optimality.
O(E log V) Heuristic Informed search
Bidirectional
Simultaneously expands from source and destination, meeting in the middle — cutting search space roughly in half on dense networks.
O(b^(d/2)) Bi-directional Optimised
Top K Routes
Yen's K-Shortest Paths returns the K best route alternatives ranked by cost or time. Gives passengers real booking options beyond a single answer.
O(K · V · (V+E) log V) K-shortest Multi-path
Reachability (BFS)
Breadth-First Search finds all airports reachable within K connections. Useful for frequent flyer coverage and connectivity analysis.
O(V + E) BFS Traversal
Budget Mode
Modified Dijkstra pruning at a cost ceiling. Returns every reachable destination within your budget in any selected currency.
O((V+E) log V) Constrained Single-source
Critical Airports
Tarjan's articulation point algorithm. Finds airports whose removal disconnects the network — single points of failure like PTY (Panama City).
O(V + E) Tarjan Graph structure
Connected Components
Tarjan's SCC algorithm. With one-way routes in this dataset (Caucasus, Central Asia, Caribbean) you will see multiple distinct strongly connected components.
O(V + E) Tarjan SCC Directed
Min Spanning Tree
Kruskal's MST with union-by-rank. Isolated sub-graphs (South Cone, Central Asia, Pacific) each form separate trees in the spanning forest.
O(E log E) Kruskal Spanning forest
Interactive Tool

Run the optimizer.

Currency
Cheapest Route
Dijkstra — minimum cost path
O((V+E) log V)

Configure the parameters above and click Run Algorithm to see results.

Technical Overview

Built on proven foundations.

Adjacency List Graph
The flight network is stored as a weighted directed adjacency list. Each edge carries both a cost and a duration weight, supporting multi-criteria optimisation without graph duplication.
Geographic Coordinates
Every airport carries latitude and longitude data. This powers the A* Haversine heuristic, the live map canvas, and the automatic path zoom feature.
Input Validation
The Python backend validates all CSV rows for missing columns, non-numeric values, and negative weights. Bellman-Ford additionally detects negative-weight cycles before returning results.