Types of Variables in Stata
Stata supports the storage of several types of variables. The broader types of variables are integers, reals, and strings.
Examples of integer variables are dummy variables (usually represented by 0 or 1), or categorical variables represented by numbers (example: 1, 2, and 3 for single, married, divorced). Integer variables, as its namesake suggests, can only take on integer values (or whole numbers for the uninitiated). Stata classifies integer variables as either byte, int, or long. Byte is used for ± one or two-digit integers, int supports up to ±32,740, and long is for integer values up to ±2.14 billion.
Real types of variables are numbers that have support for storage of decimals. The two real variable-types in Stata are float and double. Float is the default storage type when you input a variable with number entries in Stata. Float is precise up to seven digits after the decimal point while double is more precise up to the fifteenth digit.
The last type is the string variable. This type of value storage supports alphanumeric characters and can support the storage of up to 244 characters per observation. This is used in storage of information such as names.
Reference: "An Introduction to Modern Econometrics Using Stata" by Christopher Baum