Programming for Machine LearningCycle I · 2025-2
Data analysis through public APIs: statistical visualization and result export
Pulling SpaceX's public launch API to look for patterns in mission success, booster reuse, and how the company's operations changed shape over two decades.
The goal was to take a public API end to end: fetch, clean, analyse, visualize and export — with SpaceX’s historical launch data as the subject. It is a good dataset for the exercise because the interesting questions are about trend and ratio rather than single numbers.
Across 187 launches, 181 were successful — a 96.8% overall success rate. 115 of them (61.5%) flew on a reused booster, and 143 of 156 landing attempts succeeded. The charts trace launches per year and the reuse curve over time. The shift from expendable to reused hardware shows up as a slow crossover rather than a single jump.
The project also covers the unglamorous half of the work — handling the API’s response shapes, building the statistical summaries, and exporting results in a form someone else can pick up.