Apache Spark Certification Practice Test

RDD transformations being labeled as "lazy" means that these transformations are not computed immediately when they are defined. Instead, they are stored as a lineage of operations to be performed on the data when an action is called. This allows Spark to optimize the execution plan by combining multiple transformations into a single operation, minimizing resource usage and improving performance.

When transformations are defined, Spark merely remembers the operations to perform later, but the actual computations do not occur until an action, such as `collect()` or `count()`, is invoked. This characteristic is crucial for efficient computation in Spark, as it allows the system to delay processing until it can evaluate the most efficient method to execute the transformations based on the data's current state.

Thus, while the other options might suggest immediate execution or changes in memory usage, they do not reflect the true nature of laziness in the context of RDD transformations, which is fundamentally about deferring computation until necessary.