We can use ad hoc analysis to learn about important properties of the data. Some of the issues that can be easily solved with the data are:
- Statistical properties, such as mean, median, the range for numerical data
- Distinct values for numerical as well as non-numerical data
- The frequency of data occurrence
We can ask these questions on a sample of data or an entire dataset. With a distributed framework, such as Spark, it is quite easy and convenient to get answers to these questions. In fact, many of these frameworks have a simple API to support this. Ad hoc analysis can also be performed on the very raw data itself. In this case, some of the data transformations are applied as part of the process. The main purpose of the ad hoc analysis is to gain a quick understanding of some of the properties of the data.
We will use Spark to perform some hands...