Module 1 · Foundations

Module 1: construction feedback

Try practice before reading the worked notebook. Download solutions or open notebooks/practice/module01-solutions.ipynb in the complete source tree. If a download link does not open, use the offline access route.

route

Use an inclusive boundary; do not return the priority itself. For route(2), the expected result is False.

def route(priority, threshold=3):
    return priority >= threshold

partition_role

The final test must not choose a model. For partition_role('train'), the expected result is 'fit'.

def partition_role(name):
    return {"train":"fit","validation":"choose","test":"evaluate"}[name]

error_counts

Count errors by their consequences, not only total accuracy. For error_counts([False, True, True], [True, False, True]), the expected result is [1, 1].

def error_counts(actual, predicted):
    return [sum((not a) and p for a,p in zip(actual,predicted)), sum(a and not p for a,p in zip(actual,predicted))]

Complete-study results feedback

80 gives Within the chosen threshold (80 > 80 is false); 81 and 100 give Review storage; 100.5 is rejected with "storage_percent must be a finite number from 0 through 100". The rule decides only valid inputs; the input check runs first and stops impossible percentages. These values come from running the script during the October 28, 2026 review.

Reasoning task feedback

For example, [False, False, False] gives [0, 1] (one miss) and [True, True, False] gives [1, 0] (one false alert); both are 2/3 accurate. The preferred list depends on which error costs more. A missed storage problem can cause an outage, while a false alert costs a review. Accuracy alone cannot express that choice.

Judge the explanation

Compare your explanation with the criteria in the study workflow: passing the visible checks alone is developing; a changed case that you predict, implement and explain meets the task.

Transfer feedback

Frame an escalation aid with a human fallback. Explain one field unavailable at ticket arrival. A defensible response names the decision time (ticket arrival), the affected stakeholder, the evidence source and the human fallback. It also names a field that only exists after escalation, such as the resolver group.