Rescorla-Wagner (1972) error-driven associative model

Usage

rescorla_wagner(C, alpha, beta, lambda)

Arguments

C: Decay
alpha: Salience (fix at 1 unless manipulated)
beta: Learning rate – a proportion of lambda
lambda: Maximum associative value that a CS can achieve - should be larger than learning rate

Value

An object of class xslMod

Examples

mod <- rescorla_wagner(C = 1, alpha = 1, beta = 0.1, lambda = 3)
xsl_run(mod, get_example_ambiguous_condition())
#> $fits
#> $fits[[1]]
#> $fits[[1]]$sims
#> NULL
#> 
#> $fits[[1]]$responses
#>      [,1] [,2] [,3] [,4]
#> [1,]  0.5  0.5  0.5  0.5
#> 
#> $fits[[1]]$perf
#>   1   2   3   4 
#> 0.5 0.5 0.5 0.5 
#> 
#> $fits[[1]]$matrix
#>     1   2   3   4
#> 1 0.9 0.9 0.0 0.0
#> 2 0.9 0.9 0.0 0.0
#> 3 0.0 0.0 0.9 0.9
#> 4 0.0 0.0 0.9 0.9
#> 
#> $fits[[1]]$sse
#> [1] 0
#> 
#> $fits[[1]]$data
#> xslData object with label "example condition" and condition "ambiguous"
#>   training trials: 2
#>       test trials: 0
#>             words: 4
#>           objects: 4
#>        accuracies: 4
#> 
#> 
#> 
#> $sse
#> [1] 0
#> 
#> $unweighted_sse
#> [1] 0
#>