A company uses Amazon SageMaker for its ML workloads. The company's ML engineer receives a 50 MB Apache Parquet data file to build a fraud detection model. The file includes several correlated columns

Question

Options

• ADownload the file to a local workstation. Perform one-hot encoding by using a custom Python
• BCreate an Apache Spark job that uses a custom processing script on Amazon EMR.
• CCreate a SageMaker processing job by calling the SageMaker Python SDK.
• DCreate a data flow in SageMaker Data Wrangler. Configure a transform step.

• A
  2% (1)
• B
  6% (3)
• C
  9% (5)
• D
  83% (44)

Explanation

D is correct because SageMaker Data Wrangler is purpose-built for low-code/no-code data preparation - dropping columns is a built-in transform requiring only a few clicks in a visual interface, making it the lowest-effort option for a 50 MB file.

A is wrong because downloading the file locally and writing one-hot encoding (a feature encoding technique, not column dropping) is both the wrong tool for the job and unnecessary manual effort. B is wrong because spinning up an EMR cluster with a custom Spark job is heavy infrastructure overhead for a trivially small 50 MB file - EMR is designed for large-scale distributed processing. C is wrong because a SageMaker Processing Job requires writing and deploying a custom script, which is more engineering effort than a no-code visual tool.

Memory tip: Think of Data Wrangler as the "drag-and-drop spreadsheet editor" of SageMaker - whenever a question asks for least effort data transformation on a reasonably sized file, Data Wrangler beats scripted or infrastructure-heavy alternatives every time.

No community discussion yet for this question.