7. Variables

Combining objects that do not change is not very interesting task. The GNA was designed in order to work with models with large number of parameters. In the simplest case the variables represent weights in a sum of arrays (WeightedSum). In the more complex case variables represent some physical parameters, like parameters of energy resolution model (EnergyResolution).

We will start working with variables by introducing the concept of environment and namespaces.

7.1. Environment

Environment is a global GNA object defined in Python and holding a folder like structure with variables. When GNAObject is created it requests a list of variables it depends on from the environment. The variables are then bound to the transformations. As the value of the variable changed the taintflag=false is propagated to the dependent variables and transformations invalidating the cached data.

• 7.1.1. Defining parameters
  • 7.1.1.1. Gaussian parameters
  • 7.1.1.2. Nested namespaces
  • 7.1.1.3. Changing the value of the parameters and working with angles
• 7.1.2. Correlations between variables (TBD)
• 7.1.3. WeightedSum transformation
• 7.1.4. Working with nested namespaces