Practical Big Data Analytics

A technical understanding of MapReduce

MapReduce, as the name implies, has a map phase and a reduce phase. A map phase is generally a function that is applied on each element of its input, thus modifying its original value.

MapReduce generates key-value pairs as output.

Key-value: A key-value pair establishes a relationship. For example, if John is 20 years old, a simple key-value pair could be (John, 20). In MapReduce, the map operation produces such key-value pairs that have an entity and the value assigned to the entity.

In practice, map functions can be complex and involve advanced functionalities.

The reduce phase takes the key-value input from the map function and performs a summarization operation. For example, consider the output ...