# SCONE Parameters

SCONE can optimize both Control parameters and Model parameters. During optimization, parameter values are stored in a .par file. In a SCONE Scenario, parameters are specified as follows:

parameter = mean~stdev<min,max>

where `mean`

is the initial mean value, `stdev`

is the initial standard deviation, and `min`

and `max`

represent the allowed range for the parameter. A higher version for `stdev`

means the initial guesses are more spread out, which can lead to better results quickly, but also leads to higher variance between optimizations.

Alternatively, there is the following shorthand notation:

parameter = ~mean<min,max>

in which the standard deviation is set to 10% of the absolute value of mean.

### Control Parameters

The most common type of parameters to optimize are for SCONE Controllers. Examples of control parameters are:

- Data points of a piece-wise Function in a FeedForwardController.
- Reflex gains or offsets in a ReflexController.
- State transition times in a SequentialController.

### Model Parameters

Model parameters are specified in the `Model { Parameters { … } }`

section of a scenario. Examples are:

- Muscle maximum isometric force
- Muscle optimal fiber length
- Body mass or inertia
- Coordinate limit forces of a model joint