Intermediate · 30 min · 4 steps
Why averaging percentages hid a cab cancellation problem
Aniket Joshi, Fleet Operations Telemetry Engineer at MetroCab
Aniket Joshi is an engineer at MetroCab, a cab app with over 150,000 drivers in Mumbai, Bengaluru, Hyderabad and Delhi. At 6:30 PM on a Friday, the Chief Operating Officer calls an emergency meeting.
The question
How do we work out the true cancellation rate for a group of drivers in Python, so a few rogue drivers are not hidden by many honest ones?
- Weighted averages
- Simpson's paradox avoidance
- Rolling window aggregation
- Anomaly alerting
The rest of this case study is locked
The data, every step with its code and result, and the final answer open with a pass. A few case studies stay free so you can see how they work.