Which advantage do dataframes have over RDDs?

DataFrames provide built-in optimizations for queries, which is one of their key advantages over Resilient Distributed Datasets (RDDs). The Catalyst optimizer, which is part of the Spark SQL engine, applies various optimization techniques to improve query execution plans for DataFrames. This includes logical optimization, physical planning, and the use of technique such as predicate pushdown and columnar storage, which help to minimize data shuffling and read.

These optimizations enable DataFrames to execute SQL-style queries more efficiently than RDDs, which do not have a similar level of optimization. RDDs are more focused on raw data manipulation, requiring developers to manually handle performance improvements, while DataFrames abstract these complexities, allowing users to benefit from improved performance without extensive optimization efforts on their part.

Other options like flexibility in data types and compatibility with Spark Streaming, while relevant, do not fundamentally capture the core efficiency gains provided by the optimizations associated with DataFrames.