A useful handoff starts with a clear prediction question and a reliable pair of input and outcome objects. This practice set mixes those two decisions on fresh fictional records. Some teams need a number, others need a label, and one research team has no labeled outcome at all. Read the requested output before writing code.
The feature tasks require exact row correspondence. Neither an identifier nor the target belongs in the input matrix unless the task explicitly requests it. You will also see an audit that retains a record number on purpose. That exception does not make record numbers useful predictors. Keep each answer tied to the stated contract. There are no new modeling methods here, so use the earlier worked examples if the notebook syntax slows you down.