Advanced Python

The model that explains the surprises.

• Objects, references and mutability hands on
  Why a default argument of [] is a bug, and why two variables can change together.
• The data model hands on
  Dunder methods, and making your own objects behave like built-in ones.
• Iterators and generators hands on
  Processing a file larger than memory, one line at a time.
• Decorators and context managers hands on
  Written from scratch, so they stop being syntax you copy.

Code that a second person can work on.

• Packages, imports and project layout hands on
  Why the circular import happened and how the layout prevents it.
• Type hints and static checking hands on
  mypy, and catching a class of bug before running anything.
• Dataclasses and Pydantic hands on
  Validating data at the boundary rather than trusting it inward.
• Configuration and environments
  One codebase, several deployments, no secrets in git.

Beyond a handful of assertions.

• Fixtures, parametrisation and factories hands on
  Tests that stay readable as they multiply.
• Mocking, and its dangers hands on
  A mock that no longer matches reality is worse than no test.
• Coverage, and what it does not tell you
  100% coverage of code that asserts nothing.
• Testing the database and the API layer hands on
  Integration tests that catch what unit tests cannot.

Doing several things at once, correctly.

• Threads, processes and the GIL
  What is actually parallel in Python and what is not.
• asyncio hands on
  Event loops, coroutines, and where async genuinely helps.
• Concurrent I/O in practice hands on
  Fetching 500 URLs without melting the target server.
• Race conditions, demonstrated hands on
  Written, observed, then fixed — the only way this lesson lands.

Measuring before optimising.

• Profiling hands on
  Finding the actual hot path instead of guessing at it.
• Data structures and complexity in practice hands on
  The list-versus-set change that turns minutes into milliseconds.
• Caching, and invalidating it hands on
  Where a cache helps and how it silently serves stale answers.
• When to stop
  Readable code that is fast enough beats clever code that is faster.

Getting it running somewhere and keeping it there.

• Building and publishing a package hands on
  pyproject.toml, versioning, and a library a colleague can install.
• Docker hands on
  A container that runs the same on your machine and the server.
• CI that runs the tests hands on
  A pipeline that refuses a broken merge.
• Logging, metrics and the 3am question hands on
  Instrumenting so a failure can be diagnosed without you.

Before you start

• Python Programming, or equivalent
  You must already be writing working Python. This course starts where that one ends.

Where it leads

• Machine Learning
  Python is the prerequisite; this is the modelling route.
• Full Stack Web Development (Python)
  The web application route.

Roles this prepares you for

• Python Developer
  Backend services and internal tooling.
• Backend Engineer
  APIs and data pipelines.
• DevOps-leaning Developer
  For those who took to the containers and CI module.

Interested in this programme?

Try the subject free first — there is no reason to pay before you know you like it.