Instructor Guide¶

Student reports expose evidence but intentionally omit the automatically inferred bottleneck. Instructor code may inspect:

result.bottleneck_analysis()
result.likely_bottleneck
result.export_instructor_results("instructor/expected_results/lab.json")

The analyzer combines utilization, average queue accumulation, and queue waiting rather than selecting only the most utilized Machine.

Student and instructor notebooks use the same fixed seed for instructional comparisons. Deterministic exercises should provide exact expected values; stochastic exercises should provide expected ranges, replication means, sample standard deviations, and 95% confidence intervals.

Before release, execute every notebook with Run All from a fresh environment. Confirm that it creates output directories, generates its data internally, completes the simulation, renders plots and animations, writes MP4/GIF/CSV outputs, and uses no hidden notebook state. Clear generated cell outputs before committing notebooks.

Instructor notebooks should retain completed scenario code, KPI expectations, a bottleneck interpretation, common mistakes, plots, and export steps. Machine-readable expected result files live under instructor/expected_results/.