Python classes online

What went wrong

In Python, 0, empty text and empty lists all count as false. The shortcut threw away 10,478 real zeros, so the average described only the busy months and came out 55% too high.

A second, quieter error

Some 222 areas approved no new houses in any of the 12 months. With the shortcut, those areas looked as if they had no data, when the data said something definite: zero.

The simple baseline wins

Because the download has no blanks, the plainest code, with no filter at all, gives the right answer. The "clever" filter was the only thing that broke it. We teach learners to look at the data before deciding what to skip.

Why this suits every age

A twelve year old can see the problem with a list like [3, 0, 5, 0]. An adult meets it in every spreadsheet with a column of counts. The lesson is the same: decide what "missing" means, then write it exactly.

It also introduces None, one of the ideas that separates writing Python from guessing at it.

What we did not do

We report an average per area and month, our own calculation, and do not add the areas up into a national figure; the ABS publishes its own totals, and parts summed from a download need not match them.

ABS material is licensed under Creative Commons Attribution 4.0, so learners can reuse it in their own projects with credit.

For children

The habits arrive as games: "what happens if the list is empty?" becomes a challenge, not a lecture.

For teenagers

Tests are written before code, and projects are kept under version control from the second term.

For adults

Every script for real work keeps the original data untouched and prints counts before and after.