To process data it is not enough just to obtain them. You need to convert it to the appropriate format, typically to numbers. Since Galileo Galilei, who urged to “measure what can be measured, and make measurable what cannot be measured”, European science aggregated tremendous experience in transferring surrounding events into numbers. Most of our instruments are devices which translate environment features (e.g., temperature, distance) to the numerical language.
- 3.3: Colors, Names and Sexes - Nominal Data
- Nominal, or categorical, data, unlike ranked, are impossible to order or align. They are even farther away from numbers. For example, if we assign numerical values to males and females (say, “1” and “2”), it would not imply that one sex is somehow “larger” then the other. An intermediate value (like “1.5”) is also hard to imagine. Consequently, nominal indices may be labeled with any letters, words or special characters—it does not matter.
- 3.8: Inside R
- Vectors in numeric, logical or character modes and factors are enough to represent simple data. However, if the data is structured and/or variable, there is frequently a need for more complicated R objects: matrices, lists and data frames.