Add lesson on parallelisation

[JostMigenda]

This could give a brief overview over ProcessPoolExecutor/InterpreterPoolExecutor /ThreadPoolExecutor, the GIL and free-threading, async/await & a flowchart of which to use when.

See RSE-Sheffield#60 for previous discussion.

[JostMigenda]

See also RSE-Sheffield#72 for previous discussion on Amdahl’s law (which could fit into the parallelisation section), and a draft of a callout box.

[JostMigenda]

A few thoughts from #40:

• mention parallelisation external to Python (e.g., using slurm to submit job arrays to an HPC cluster)
• Linear algebra libraries underlying NumPy (BLAS/LAPACK) may be parallelised. We have a callout box on that in the NumPy lesson already; might be worth cross-referencing in the parallelisation episode
• Talk about thread safety (e.g., is it safe to parallelise code that’s writing to the same Python object or data frame?). The answer is probably just “as a first rule of thumb, reading is usually fine and writing isn’t”, but we can refer to official documentation or other resources for details. (As part of the free-threading work, I remember seeing a couple of new resources on that recently.)