# Relations

Most datasets are rectangular. There are variables (usually represented as columns) and cases (usually represented as rows). Each case has a value recorded for each variable. The recorded value may be a value defined as “missing” rather than a number or a category name. Because each case has a value for each variable and each variable has a value at each case, the array of data can be shown as a rectangular table of values.

Data analyses typically relate two or more variables to each other. However, the variables must hold data for the same cases in the same order. If a variable recording median education in each of the 50 states was arranged in alphabetical order, it would make no sense to plot it against a variable holding median income in each state that was ordered from west to east, or against a variable that recorded the heights of 50 people in a sample.

This rectangular structure is known in database theory as a relation, and Data Desk adopts this terminology. Formally, each row in a relation must be unique. Accordingly, Data Desk assigns a unique case number to each row in order from top to bottom.

If your dataset is a standard rectangular data table, calling it a relation changes nothing. However, if your data include variables recorded for several relations, you will find that Data Desk’s relational data management abilities let you structure, enter, and work with your data in more natural ways.

For most datasets, Data Desk uses relations to make your life easier automatically. For example, if your data form a simple relation, Data Desk automatically keeps cases aligned in your variables. Thus, if you cut a case out of one variable, Data Desk offers to delete that case from all variables in the relation to preserve your ability to analyze the variables together.

Most analyses that deal with more than one variable make sense only when the variables are in the same relation. You cannot combine variables from two different relations in the same plot or calculation, but Data Desk provides ways to refer from one relation to another so that the resulting variables are properly matched.

## New Relation

The **{Data} New>Relation** command creates a new relation and opens a variable window for it. You can drag variables into the relation or enter new data there.

## Relational Functions

Data Desk offers functions to work with relations. You’ll find them here.