In Fabric with Apache Spark, which artifact should you create to explore data interactively?

In Fabric with Apache Spark, which artifact should you create to explore data interactively?

• A. A Spark job definition.
• B. A Data Factory pipeline.
• C. A notebook.
• D. A machine learning model.

Answer: (C)

Interactive data exploration relies on an environment that lets you write and execute code, inspect results immediately, and iterate quickly. A notebook provides exactly that: you can run Spark-compatible code (like PySpark or Spark SQL) against a Spark pool, query data, create visualizations, and add narrative text to explain each step. This immediate feedback loop makes it ideal for exploring datasets, trying hypotheses, and documenting findings in one place. The other artifacts serve different purposes. A Spark job definition is for submitting and running batch Spark workloads, not for interactive exploration. A Data Factory pipeline orchestrates a sequence of data-processing steps, which is great for automation and workflow management but not for ad-hoc analysis. A machine learning model is about training and using predictive models, not about the hands-on, iterative data examination you do during exploration.

Interactive data exploration relies on an environment that lets you write and execute code, inspect results immediately, and iterate quickly. A notebook provides exactly that: you can run Spark-compatible code (like PySpark or Spark SQL) against a Spark pool, query data, create visualizations, and add narrative text to explain each step. This immediate feedback loop makes it ideal for exploring datasets, trying hypotheses, and documenting findings in one place.

The other artifacts serve different purposes. A Spark job definition is for submitting and running batch Spark workloads, not for interactive exploration. A Data Factory pipeline orchestrates a sequence of data-processing steps, which is great for automation and workflow management but not for ad-hoc analysis. A machine learning model is about training and using predictive models, not about the hands-on, iterative data examination you do during exploration.