Given a denormalized dataset, how would you extract the desired columns or transform it into a normalized form using Python, Spark, or SQL?

When you have a denormalized table—say a single row contains multiple comma‑separated values or nested JSON—you can transform it into a normalized form using SQL, Spark, or Python. In SQL, you can use STRING_SPLIT or CROSS APPLY to explode comma‑separated columns, then SELECT the desired fields. In Spark, you can read the data into a DataFrame, use the explode function on an array column, or parse JSON with from_json, then select and write the result. In Python, you can read the CSV with pandas, use str.split and explode, or parse JSON with json.loads, then flatten the structure. The key steps are: 1) Identify the denormalized columns; 2) Split or parse them into separate rows or columns; 3) Apply any necessary transformations; 4) Write the normalized data back to a target table or file. This approach keeps the data consistent and queryable.