Skip to content

Tutorial: Identifying top food delivery riders by highest delivery speed per distance. #43

Description

@jens-guenther

There's a zomato riders' dataset available on kaggle: https://www.kaggle.com/datasets/cbhavik/swiggyzomato-order-information

Based on that dataset create a tutorial to identify:

  • the "wormholers": riders with the highest (perc80) delivery speed per mile and day
    (- the "cheaters": riders with fast acceptance time (perc80) but slow delivery speed per mile)

The tutorial should cover:

Given: there's a kafka cluster already running (perhaps an tutorial of it's own?)
Given: the data is available at a topic (small linux commandline is fine)

  1. App Foundations: setup a micronaut Kipes app. (from(topic).logInfo(date, rider, distance))
  2. Calculating delivery speed per mile into new topic (automatic topic creation vs. SDK, incl. how to reset the event index for the consumer via cmd line)
  3. Adding perc80 (delivery speed per mile) per day
  4. Filter and Table per day

(The "cheater" addition could show how to use joins but that could get too complicated for one tutorial)

Metadata

Metadata

Assignees

Labels

No labels
No labels

Type

No type

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions