Data Entities

<< Click to Display Table of Contents >>

Navigation:  Defining Entity Types >

Data Entities

Previous pageReturn to chapter overviewNext page

Time series data from spreadsheets or databases appear in the model as data entities. Data entities contain variables whose values are read in from external sources. Typically, data entities are used to provide time-series data to the model. These data may be used as exogenous information, representing signals you don't wish to calculate from equations in the model. For static data, see Initializing Entities.

 

Once you have defined a data entity for your data, other entities may access the data by using references to the data entity. Data may also be used in calibration to measure the misfit between your model and the data.

 

Using data entities requires 3 things:

 

1. A data source (a spreadsheet or database) with time-stamped observations of one or more variables

2. A data entity, with variable names representing the variables in the data

3. A data map, which links specific variables in the data source to variables in the data entity and specifies interpolation methods.

 

It is easiest to begin with the data source. From there you can create the data map and the data entity, in either order.

 

A simple example of using historical data in a model computation appears in the Wilshire 5000 application.

 

Note that data entities provide external information to the model during the course of simulation. This is distinct from initialization data, which tell the model how many of each entity to create, with what characteristics at the start of simulation or at the time new entities are introduced. Initialization is covered in the Initializing Entities section.