Print so you can see what happened

The print line is your event log for this lab. Structured text beats “it worked :)” because you can compare it to the expected evidence character by character.

truth = [1, 0, 1, 1, 0]
scores = [0.8, 0.3, 0.6, 0.9, 0.2]
threshold = 0.5
preds = [int(score >= threshold) for score in scores]

pairs = list(zip(truth, preds))
tp = pairs.count((1, 1)); fp = pairs.count((0, 1))
tn = pairs.count((0, 0)); fn = pairs.count((1, 0))
accuracy = (tp + tn) / len(pairs)
print(preds, f"accuracy={accuracy:.2f}", tp, fp, tn, fn)

Expected evidence:

[1, 0, 1, 1, 0] accuracy=1.00 3 0 2 0

Design a tiny checklist beside the print:

• threshold — was it still 0.5?
• preds — do guesses match hand counting?
• tp fp tn fn — do they sum to 5?
• accuracy — does it match (tp+tn)/5?
• label the run fixture-only so nobody mistakes it for a huge real dataset.

Turn one past bug (mismatched lengths, threshold out of range) into a test that fails if the warning disappears. Observability without a failing check is decoration; with a check, it is part of the toy.

Do not print real inbox text, phone numbers, or passwords. Print scores, truths, and counts. If you demo “spam message?” use made-up strings like "FREE PRIZE!!!" — never a friend’s real chat.

Imagine a teammate opens only your printouts after something went weird. They should be able to tell apart:

1. Wrong lists — lengths mismatched or empty, so the scoreboard never should have run.
2. Wrong cutoff — threshold drifted, so guesses flipped even though truths stayed put.
3. Honest fixture win — the expected evidence line appears, labeled practice-only.

If those three stories look identical in your notes, rename fields until they do not. Good observability is less about more printing and more about printing the few facts that settle arguments. Add a one-line fixture_id (for example five-round-spam-v1) whenever you change a truth or score so old screenshots do not confuse new runs.