Glossary

Attribute

A specific type of value that is associated with each simulant. For example, each simulant may have an attribute to describe their age or position. These are computed by attribute pipelines and stored in the population state table.

Attribute Pipeline

A specific type of pipeline whose callable must accept a simulant index and return tabular data of attributes for those simulants.

Component

Any self-contained piece of code that can be plugged into the simulation to add some functionality. In vivarium we typically think of components as encapsulating and managing some behavior or attributes of the simulants.

Configuration

The configuration is a set of parameters a user can modify to adjust the behavior of components. Components themselves may provide defaults for several configuration parameters.

Metrics

Simulation outputs, often used interchangeably with “results”.

Model Specification

A complete description of a vivarium model. This description details all plugins, components, and configuration required to run the model. A model specification is stored in a yaml model specification file which is parsable by the simulation framework.

Pipeline

A vivarium value pipeline. A pipeline is a framework tool that allows users to dynamically construct and share data across several components. See also: Attribute Pipeline.

Plugin

A plugin is a python class intended to add additional functionality to the core framework. Unlike a normal component which adds new behaviors and attributes to simulants, a plugin adds new services to the framework itself. Examples might include a new simulation clock, a results handling service, or a logging service.

Population State Table

The core representation of the population in the simulation. Often referred to as simply the “state table”, it consists of a row for each simulant and a column for each attribute.

Private Column

The source of a corresponding attribute pipeline. Only the component that creates a private column has read or write access to it. Note that not all attribute pipelines require private columns as a source, but any private columns created are automatically paired with an attribute pipeline of the same name.

Public Column

A column in the population state table that is generated by an attribute pipeline and is read-accessible by all components.

Simulant

An individual or agent. One member of the population being simulated.

Value

A dynamic variable in the simulation that is computed by a pipeline and can be shared across multiple components.