When should we use dataframes and datasets in spark?
Both support scala
and are more efficient than RDDs when talking about structured and semi-structured data.
However, Is it ambiguous to choose datasets rather than dataframes ? Is the only advantage of datasets over dataframes TypeSafety?