Run the working guess path

You are not shipping an app store product. You are proving: for this fixture and this threshold, guesses and confusion counts are exact. Prefer boring, readable Python over a mysterious library you cannot explain to a friend.

Run the path twice. Matching output both times is a feature — like a basketball drill that lands the same way when form is steady.

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

Take score 0.8 with truth 1. Score ≥ 0.5 → guess 1. Pair (1, 1) is a true positive (correct yes) — like calling “make” when the ball really went in.

Take score 0.3 with truth 0. Guess 0. Pair (0, 0) is a true negative (correct no).

On this fixture every guess matches: three true positives, two true negatives, zero false alarms, zero misses. Accuracy is (3 + 2) / 5 = 1.00.

School metaphor: TP is correctly spotting a spam text; FP is wrongly accusing a normal text; TN is correctly leaving a normal text alone; FN is missing a real spam text.

While you run, make sure you can narrate:

1. Each score compared to the threshold.
2. The preds list.
3. The four bucket counts and accuracy.

If you can only say “it printed something,” you do not yet own the happy path.