7. Data Representation

On one level, data are just bytes in a computer disk. But, these bytes can be very useful when we attach them to things that happen in the real world. Drivers in Los Angeles use a service called Sigalert 1 to see which freeways have jams (although in Los Angeles, the answer is often “all of them”). Sensors installed on freeway roads record the speed of cars passing by. Sigalert takes that data, makes a map, and displays it to users:


Sigalert displays data that represent traffic conditions. Data scientists navigate a few levels of data representation. The data’s structure and format refers to the way the computer represents the data. SQL databases store data in a table structure, but other kinds of data structures exist. Data semantics refer to the real-world meaning that we as data scientists assign to each datum. Every data analysis begins by understanding what the data represent.