In the Map Reduce framework, there is a need for translating the programs into a sequential Map and Reduce stages. Pig is an abstraction that supports the data analysts during the translation of the programs into a series of the stages.(Approximately 10 lines of Pig code are equal to 200 lines of MapReduce code)

It has the ability to process and analyze the large data sets, which is a scripting platform that is used with Apache Hadoop

It consists of a compiler in the infrastructure layer to produce the sequences of MapReduce programs, in which large scale parallel implementations already exist

Its language layer currently consists of a textual language called Pig Latin

We can do all data manipulations in Hadoop with Pig

It works with data from many sources, including structured and unstructured data, and store the results into the Hadoop Data File System

Extensibility

Easily programmed

Self- optimizing