Which of the following code blocks returns a DataFrame where column divisionDistinct is the approximate number of distinct values in column division from DataFrame storesDF?

Question

Options

• AstoresDF.withColumn("divisionDistinct", approx_count_distinct(col("division")))
• BstoresDF.agg(col("division").approx_count_distinct("divisionDistinct"))
• CstoresDF.agg(approx_count_distinct(col("division")).alias("divisionDistinct"))
• DstoresDF.withColumn("divisionDistinct", col("division").approx_count_distinct())
• EstoresDF.agg(col("division").approx_count_distinct().alias("divisionDistinct"))

• A
  14% (6)
• B
  7% (3)
• C
  74% (31)
• D
  2% (1)
• E
  2% (1)

Explanation

approx_count_distinct() is an aggregate function from pyspark.sql.functions. Aggregate functions must be used inside agg() or groupBy().agg(), and the result should be aliased. The correct code is storesDF.agg(approx_count_distinct(col('division')).alias('divisionDistinct')). Option C matches this exactly. Option A uses withColumn with an aggregate function, which produces incorrect behavior (same aggregated value repeated per row, not a single summary row). Options B, D, and E incorrectly try to call approx_count_distinct() as a method on a Column object, which does not exist.

No community discussion yet for this question.