Data Cleaning Checklist

1. 内容を入力する

下の値は例です。自分の内容に置き換えてください。保存・送信はされません。

2. プロンプト全文をコピー

Complete the task using these inputs. Fields: [Fields]; Problem: [Problem]. First propose a read-only inspection plan covering how duplicates are defined, missing values, date formats, and numeric anomalies. Design cleaning steps that preserve the raw data along with a change log; don't treat identical names as the same person, and output metrics to compare before and after. Structure the output as: understanding and assumptions, ready-to-use content, information I need to confirm, and next steps. Use clear, short sentences and avoid empty slogans; if key input is missing, ask at most three specific questions instead of filling gaps with guesses. Finally, check each item against my purpose and state clearly which results still need human review.

Complete the task using these inputs. Fields: [Fields]; Problem: [Problem]. First propose a read-only inspection plan covering how duplicates are defined, missing values, date formats, and numeric anomalies. Design cleaning steps that preserve the raw data along with a change log; don't treat identical names as the same person, and output metrics to compare before and after. Structure the output as: understanding and assumptions, ready-to-use content, information I need to confirm, and next steps. Use clear, short sentences and avoid empty slogans; if key input is missing, ask at most three specific questions instead of filling gaps with guesses. Finally, check each item against my purpose and state clearly which results still need human review.

3. 結果を確認する

Are the deduplication rules clear? Can the raw data be restored? Verify names, numbers, and qualifiers from the input.

満足できないときは、変数を一つずつ変えて前の結果と比べましょう。