The Crunch REST API takes a decidedly column-oriented approach to data. A “column” is simply a sequence of values of the same type. A “variable” binds a name (and other metadata) to the column, and indeed may possess a series of columns over its lifetime as inserts and updates are made to it. A “dataset” is a set of variables. Each variable in the dataset is sorted the same way; the variables together form a relation. Reading the N’th item from each variable produces a row.
Interaction with the Crunch REST API is by variables and columns. When you add data to Crunch, you send a set of columns. When you fetch data from Crunch, you send a set of variable expressions and receive a set of columns. When you update data in Crunch, you send a set of expressions which tells Crunch how to update variables with new column data.
The Crunch API consists of just a few primitive objects, arranged differently for each request and response. Learning the basic components will help you create the most complicated queries.