make attentional diffusion interface consistent

Project.toml

@@ -1,7 +1,7 @@
 name = "SequentialSamplingModels"
 uuid = "0e71a2a6-2b30-4447-8742-d083a85e82d1"
 authors = ["itsdfish"]
-version = "0.3.0"
+version = "0.3.1"
 
 [deps]
 ConcreteStructs = "2569d6c7-a4a2-43d3-a901-331e8e4be471"

README.md

@@ -8,9 +8,9 @@ The examples below show basic usage. Addition information can be found in the RE
 ```julia
 using SequentialSamplingModels
 dist = LNR(μ=[-2,-3], σ=1.0, ϕ=.3)
-data = rand(dist, 10)
-like = pdf.(dist, data)
-loglike = logpdf.(dist, data)
+choice,rts = rand(dist, 10)
+like = pdf.(dist, choice, rts)
+loglike = logpdf.(dist, choice, rts)
 ```
 
 ## Linear Ballistic Accumulator
@@ -47,9 +47,9 @@ loglike = logpdf.(dist, rt)
 ```julia
 using SequentialSamplingModels
 dist = DiffusionRace(;ν=[1.0,.5], k=0.5, A=1.0, θ=.2)
-data = rand(dist, 10)
-like = pdf.(dist, data)
-loglike = logpdf.(dist, data)
+choice,rt = rand(dist, 10)
+like = pdf.(dist, choice, rts)
+loglike = logpdf.(dist, choice, rts)
 ```
 
 ## Attentional Diffusion Model
@@ -84,7 +84,7 @@ mutable struct Transition
                     .015 .98 .005;
                     .45 .45 .1])
 
- rts = rand(model, 1000, attend, tmat)
+ choice,rts = rand(model, 1000, attend, tmat)
  ```
 
  ## Multiattribute Attentional Drift Diffusion

src/AttentionalDiffusion.jl

@@ -54,13 +54,13 @@ Generate `n_sim` simulated trials from the attention diffusion model.
 - `kwargs...`: optional keyword arguments for the `fixation` function
 """
 function rand(dist::AbstractaDDM, n_sim::Int, fixation, args...; rand_state! = _rand_state!, kwargs...)
-    rts = [Vector{Float64}() for _ in 1:2]
+    choice = fill(0, n_sim)
+    rt = fill(0.0, n_sim)
     for sim in 1:n_sim 
         rand_state!(args...; kwargs...)
-        choice,rt = _rand(dist, () -> fixation(args...; kwargs...))
-        push!(rts[choice], rt)
+        choice[sim],rt[sim] = _rand(dist, () -> fixation(args...; kwargs...))
     end
-    return rts
+    return (;choice,rt)
 end
 
 function _rand_state!(tmat)

test/attentional_diffusion_tests.jl

@@ -28,19 +28,19 @@
                             .015 .98 .005;
                             .45 .45 .1])
 
-        rts1 = rand(model, 1000, x -> attend(x), tmat)
-        n1,n2 = length.(rts1)
-        @test n1 > n2
+        choice,rts = rand(model, 1000, x -> attend(x), tmat)
+        p1 = mean(choice .== 1)
+        @test p1 > .50
 
         model = aDDM(;ν1=4.0, ν2=5.0)
-        rts2 = rand(model, 1000, x -> attend(x), tmat)
-        n1,n2 = length.(rts2)
-        @test n1 < n2
+        choice,rts1 = rand(model, 1000, x -> attend(x), tmat)
+        p1 = mean(choice .== 1)
+        @test p1 < .50
 
-        μ_rts1 = mean(vcat(rts1...))
+        μ_rts1 = mean(rts1)
         model = aDDM(;ν1=5.0, ν2=5.0)
-        rts3 = rand(model, 1000, x -> attend(x), tmat)
-        μ_rts3 = mean(vcat(rts3...))
+        choice,rts3 = rand(model, 1000, x -> attend(x), tmat)
+        μ_rts3 = mean(rts3)
         @test μ_rts1 < μ_rts3
     end
 end