Which expression sorts the grouped counts by ProductKey in descending order?

Which expression sorts the grouped counts by ProductKey in descending order?

• A. df.groupBy("ProductKey").count().orderBy("ProductKey")
• B. df.groupBy("ProductKey").count().sort("count", ascending=True)
• C. df.groupBy("ProductKey").agg({"ProductKey": "count"})
• D. df.groupBy("ProductKey").count().sort("count", ascending=False)

Answer: (D)

Sorting after a groupBy is about ordering by the result of the aggregation—the count per ProductKey—so you see the most frequent keys first. The expression that does this uses the computed count column and orders it in descending fashion. df.groupBy("ProductKey").count().sort("count", ascending=False) groups by ProductKey, counts how many records share each key, and then sorts that resulting count column from largest to smallest, giving the highest counts at the top. Why this is the best fit: after the aggregation, you want to rank by the actual counts, not by the key values themselves. Sorting by the count in descending order makes the most frequent ProductKey appear first, which is the typical goal when you want to identify top keys. Why the other approaches aren’t right: ordering by ProductKey would sort by the key values themselves, not by how many records share each key. Sorting the count in ascending order would flip the ranking, showing the least frequent keys first. And not sorting at all leaves the order unspecified, so you wouldn’t reliably see the top counts first.

Sorting after a groupBy is about ordering by the result of the aggregation—the count per ProductKey—so you see the most frequent keys first. The expression that does this uses the computed count column and orders it in descending fashion. df.groupBy("ProductKey").count().sort("count", ascending=False) groups by ProductKey, counts how many records share each key, and then sorts that resulting count column from largest to smallest, giving the highest counts at the top.

Why this is the best fit: after the aggregation, you want to rank by the actual counts, not by the key values themselves. Sorting by the count in descending order makes the most frequent ProductKey appear first, which is the typical goal when you want to identify top keys.

Why the other approaches aren’t right: ordering by ProductKey would sort by the key values themselves, not by how many records share each key. Sorting the count in ascending order would flip the ranking, showing the least frequent keys first. And not sorting at all leaves the order unspecified, so you wouldn’t reliably see the top counts first.