Automated Data Collection with R: A Practical Guide to Web Scraping and Text Mining

7 SQL and relational databases

Handling and analyzing data are key functions of R. It is capable of handling vectors, matrices, arrays, lists, data frames as well as their import and export, aggregation, transformation, subsetting, merging, appending, plotting, and, not least, analysis. If one of the standard data formats does not suffice there is always the possibility of defining new ones and incorporating them into the R data family. For example, the sp package defines a special purpose data object to handle spatial data (see Bivand and Lewin-Koh 2013; Pebesma and Bivand 2005). So, why should we care about databases and yet another language called SQL?

Simple and everyday processes like shopping online, browsing through library catalogs, wiring money, or even buying sweets in the supermarket all involve databases. We hardly ever realize that databases play an important role because we neither see nor interact with them directly—databases like to work behind the scenes. Whenever data are key to a project, web administrators will rely on databases because of their reliability, efficiency, multiuser access, virtually unlimited data size, and remote access capabilities.

Regarding automated data collection, databases are of interest for two reasons: First, we might occasionally get direct access to a database and should be able to cope with it. Second, and more importantly, we can use databases as a tool for storing and managing data. Although R has a lot of useful data management ...