class.DOR: Delay of Reinforcement Gradient (DOR) class

In Don-Li/CAB_original:

Description

Delay of Reinforcement Gradients are represented as elements in the DOR class. The use of DORs can be found in Catania (2005) and Berg & McDowell (2011). The DOR class is used to set up the DOR function, that is, its parameters are not set here. Instead, the DOR will go into a DOR_control object that contains both the DOR as well as its associted parameters.

Usage

DOR.show

remove.DOR.show()

Format

The show method prints the function that is contained in the DOR object as well as the type of object (i.e. "elemental") and the class (i.e. "DOR" ).

Details

In the CAB package, objects from classes like DOR are called 'elemental' objects That is, they comprise a particular element of the simulation without any specified parameter values. 'Control' objects are elemental objects with an associated list of parameter values.

The way to put a DOR in your model is as follows:

1. Code up the function that you want to use for the DOR or use an existing function.

2. Construct a DOR object from the function.

3. Make a list specifying the values of the parameters for the DOR function.

4. Make a DOR_control object, which contains the DOR object and the parameters. See make.control.

The show method for the DOR class can be removed with the function remove.DOR.show() and reinstated with DOR.show.

Slots

DOR

This will contain a DOR function. See examples.

name

The name of the DOR object. Inherited from elemental.

type

This will be elemental because the DOR is an elemental object. Inherited from elemental.

Built-in DOR functions

Here is a list of the built-in DOR functions. The function definitions can be seen by calling the name of the function without brackets. For each of these functions, t is the time difference between a response and the latest reinforcement. Each function takes three parameters, t, max as the maximum reinforcement effect of an individual behaviour, and scale as a scaling parameter of the DOR.

linear_DOR_fx

A DOR where the reinforcement effect is a linear function of time since latest reinforcement.

f(t) = max - 1/scale * t

Note that the reinforcemet effect is bounded by zero.

exponential_DOR_fx

A DOR where the reinforcement effect is an exponential function of the time since latest reinforcement.

f(t) = max * exp( -t/scale )

reciprocal_DOR_fx

A DOR where the reinforcement effect is a reciprocal function of the time since latest reinforcement.

f(t) = max / ( t+1 )^( 1/scale )

hyperbolic_DOR_fx

A DOR where the reinforcement effect is a hyperbolic function of the time since latest reinforcement.

f(t) = 2 * max / ( ( t + 1 )^( 1/scale ) + 1 )

With respect to the actual implementation of these DOR functions, each of the built-in DOR functions have five arguments.

max

The maximum increment from an individual response.

scale

A scaling factor on the DOR. The scaling is independent of max.

IRI_resp_times

A vector containing response times counting backward from the most recent reinforcement time.

reserve

A numeric specifying the value of the reserve.

Each of these arguments can instead be a string specifying the variable that should be taken as the argument from a input object.

The DOR function will return the value of the reserve after the increment.

Make a DOR object from a DOR function

Use the make.DOR function to make a DOR object.

Usage

make.DOR(DOR_fx, name)

Arguments

DOR_fx

A function that represents the DOR of interest

name

A character vector specifying the name of your DOR object

Value

Returns a DOR object.

References

Berg, J. P., & McDowell, J. J (2011). Quantitative, steady-state properties of Catania's computational model of the operant reserve. Behavioural Processes, 87(1), 71-83. https://doi.org/10.1016/j.beproc.2011.01.006

Catania, A. C. (2005). The operant reserve: A computer simulation in (accelerated) real time. Behavioural Processes, 69(2), 257-278. https://doi.org/10.1016/j.beproc.2005.02.009

See Also

make.control for making DOR_control objects.

set.custom_elemental for setting custom elemental classes.

make.custom_elemental for making custom elemental objects.

e_show for the function that is called by the show method.

accessor_helpers for the helper functions used in the built-in DOR functions.

class.elemental for the parent elemental class.

get.reserve for a helper function that gets the reserve value.

get.IRI_resp_times for a helper function that gets the inter-reinforcement interval response times.

Examples

# Look at the definition of the linear DOR
linear_DOR_fx
# Look at the form of the DOR
iri_resp_times = seq(0, 15 , by = 0.1 )
my_DOR = linear_DOR_fx( max = 1, scale = 10, iri_resp_times, reserve_value = 0 )
plot( iri_resp_times, my_DOR, ylab = "increment to the reserve", xlab = "time since last rft" )
# Make a DOR object out of the linear DOR function
linear_DOR = make.DOR( linear_DOR_fx, "linear_DOR" )
# Look at the form of the DOR with the "DOR" object
my_DOR2 = linear_DOR@DOR( max = 1, scale = 10, iri_resp_times, reserve_value = 0, ylab = "increment to the reserve", xlab = "time since last rft" )
plot( iri_resp_times, my_DOR2 )

# Look at the exponential DOR
exponential_DOR_fx
my_expo_DOR = exponential_DOR_fx( max = 1, scale = 10, iri_resp_times, reserve_value = 0 )
plot( iri_resp_times, my_expo_DOR, ylab = "increment to the reserve", xlab = "time since last rft" )

Don-Li/CAB_original documentation built on May 6, 2019, 2:53 p.m.