What is a delayed operation in Spark?

Delayed Operations in Spark refer to transformation operations that are not executed immediately, but rather wait until an action operation is triggered. This delayed execution helps Spark optimize execution plans and improve performance. Common delayed operations include map, filter, reduce, and other transformation operations, while action operations include collect, count, save, etc. By delaying operations, Spark can build a complete execution plan graph based on requirements, and then execute it once to avoid frequent data transfers and calculations.