The basic class for dealing with data is called Dataset,
implemented in
opus_core.datasets.dataset.
A dataset is a collection of attributes for a particular type of entity, such as a set of grid cells, or a set of households. Each member in this set has the same set of characteristics, such as income of households. In Opus, these characteristics are called attributes.
Conceptually, a dataset is similar to a table. Each attribute is a column. Each row describes the attribute values for one member of the dataset.
Some attributes are read from a data store, we call them primary attributes. Others, computed attributes, are computed by Opus variable definitions. Attributes can be also modified or created by models.