Sliding Window Aggregation in Python | Dataford Interview Questions

Lyft receives a chronological stream of completed ride events. For each event, report the number of rides and total fare for that city during the trailing time window ending at the event timestamp.

An event is represented as [timestamp, city, fare]. The window is left-exclusive and right-inclusive: for timestamp t and window size w, include events with timestamp > t - w and timestamp <= t. Each event produces one output record after it is added to the window.

Formal Specification

Implement aggregate_ride_metrics(events, window_seconds).

events is a list of [int timestamp, string city, number fare] entries sorted by nondecreasing timestamp.
window_seconds is a positive integer.
Return a list of dictionaries in input order. Each dictionary must contain timestamp, city, ride_count, and total_fare.
total_fare should preserve the numeric type resulting from adding the input fares.

Constraints

1 <= len(events) <= 10^5
1 <= window_seconds <= 10^9
Timestamps are nonnegative and sorted in nondecreasing order
City names are nonempty strings
Fares are nonnegative integers or floating-point values

Function Signature

def aggregate_ride_metrics(events, window_seconds):

Your question is Sliding Window Aggregation in Python. Start with the requirements in the Question tab.

Run and submit as often as you like. When you're ready, talk me through your approach or go straight to the code.

Problem