Skip to content

Latest commit

 

History

History
46 lines (32 loc) · 1.64 KB

README.md

File metadata and controls

46 lines (32 loc) · 1.64 KB

Intergalactic pair programming brief

We'll be your copilot(s) during this pair programming exercise, and we'll be looking at your thought process, research skills and Python proficiency.

BE OVERLY VERBOSE/TALKATIVE. Run yourself with --verbose e.g. ask questions or explain your rationale. That'll give us better insights into your capabilities, and let us hand you gists/code snippets we've got at the ready so you don't have to do too much research.

Context

The business has ingested a lot of Star Wars data in its Data Warehouse as a DuckDB database. It now has great needs for insights, and all the BI Analysts are on holiday.

Setup

A DuckDB database file is provided to you in this repo. To install DuckDB you can follow along at: https://duckdb.org/install/. You can then run the local DuckDB UI and connect to the database file in a terminal:

$ duckdb -ui ./wh.db

Navigate in your web browser to: http://localhost:4213/.

Not supported specifically on Windows machines with an ARM64 architecture.

Alternatively, try with docker:

$ docker run --rm -it -v "$(pwd):/workspace" -w /workspace --net host duckdb/duckdb duckdb -ui wh.db

Ask

Answer one or more from the following questions:

  • Show each character name along with each vehicle name this character drives (Easy)
  • Show a count of each gender that each movie includes; show the movie name, order by release date (Med)
  • Show the most prolific pair of characters i.e. show pairs of character names and count how many times they appear together accross movies (Hard)

Hints:

  • Keep an eye out for discrepancies and missing values!
  • Think of organising your SQL in CTEs