clean up

.gitignore

@@ -1 +1,2 @@
-/Manifest.toml
+*Manifest.toml
+temp.jl

Project.toml

@@ -1,7 +1,7 @@
 name = "AdaptiveDesignOptimization"
 uuid = "5f210a59-4c6a-4c2f-87d5-5cd98d51789a"
 authors = ["itsdfish"]
-version = "0.1.3"
+version = "0.1.4"
 
 [deps]
 Distributions = "31c24e10-a181-5473-b8eb-7969acd0382f"

examples/delay_discounting/Project.toml

@@ -0,0 +1,17 @@
+[deps]
+AdaptiveDesignOptimization = "5f210a59-4c6a-4c2f-87d5-5cd98d51789a"
+DataFrames = "a93c6f00-e57d-5684-b7b6-d8193f3e46c0"
+Distributions = "31c24e10-a181-5473-b8eb-7969acd0382f"
+Pigeons = "0eb8d820-af6a-4919-95ae-11206f830c31"
+Random = "9a3f8284-a2c9-5f02-9a11-845980a1fd5c"
+Revise = "295af30f-e4ad-537b-8983-00126c2a3abe"
+StatsFuns = "4c63d2b9-4356-54db-8cca-17b64c39e42c"
+StatsPlots = "f3b207a7-027a-5e70-b257-86293d7955fd"
+Turing = "fce5fe82-541a-59a6-adf8-730c64b5f9a0"
+
+[compat]
+DataFrames = "1"
+Distributions = "0.25.0"
+Random = "1"
+Revise = "3"
+StatsPlots = "0.15.0"

Examples/Delay_Discounting/Run_Delay_Discounting.jl → examples/delay_discounting/Run_Delay_Discounting.jl

@@ -6,19 +6,24 @@ cd(@__DIR__)
 # load the package manager
 using Pkg
 # activate the project environment
-Pkg.activate("../../")
-using Revise, AdaptiveDesignOptimization, Random, Distributions
+Pkg.activate("")
+using Revise
+using AdaptiveDesignOptimization
+using Distributions
+using Random
 includet("Delay_Discounting.jl")
 #######################################################################################
 #                               Define Experiment Design
 #######################################################################################
 Random.seed!(120341)
 prior = [Uniform(-5, 5), Uniform(-5, 50)]
 
-model = Model(;prior, loglike)
+model = Model(; prior, loglike)
 
-parm_list = (κ = range(-5, 0, length=50) .|> x->10^x, 
-   τ = range(0, 5, length=11)[2:end])
+parm_list = (
+    κ = range(-5, 0, length=50) .|> x->10^x, 
+    τ = range(0, 5, length=11)[2:end]
+)
 
 # parm_list = (κ = [.1], 
 #    τ = [.2,.5])

examples/monetary_gambles/Project.toml

@@ -0,0 +1,16 @@
+[deps]
+AdaptiveDesignOptimization = "5f210a59-4c6a-4c2f-87d5-5cd98d51789a"
+DataFrames = "a93c6f00-e57d-5684-b7b6-d8193f3e46c0"
+Distributions = "31c24e10-a181-5473-b8eb-7969acd0382f"
+Random = "9a3f8284-a2c9-5f02-9a11-845980a1fd5c"
+Revise = "295af30f-e4ad-537b-8983-00126c2a3abe"
+StatsPlots = "f3b207a7-027a-5e70-b257-86293d7955fd"
+UtilityModels = "bddc258c-9485-48da-848b-3e4c1ee8966c"
+
+[compat]
+DataFrames = "1"
+Distributions = "0.25.0"
+Revise = "3"
+StatsPlots = "0.15.0"
+UtilityModels = "0.3.0"
+Random = "1"

Examples/Monetary_Gambles/Run_Monetary_Gamble.jl → examples/monetary_gambles/Run_Monetary_Gamble.jl

@@ -6,8 +6,12 @@ cd(@__DIR__)
 # load the package manager
 using Pkg
 # activate the project environment
-Pkg.activate("../../")
-using Revise, AdaptiveDesignOptimization, Random, UtilityModels, Distributions
+Pkg.activate("")
+using Revise
+using AdaptiveDesignOptimization
+using Distributions
+using Random
+using UtilityModels
 includet("TAX_Model.jl")
 Random.seed!(25974)
 #######################################################################################
@@ -26,8 +30,8 @@ parm_list = (
 dist = Normal(0,10)
 n_vals = 3
 n_choices = 2
-design_vals = map(x->random_design(dist, n_vals, n_choices), 1:1000)
-filter!(x->abs_zscore(x) ≤ .4, design_vals)
+design_vals = map(x -> random_design(dist, n_vals, n_choices), 1:1000)
+filter!(x -> abs_zscore(x) ≤ .4, design_vals)
 design_names = (:p1,:v1,:p2,:v2)
 design_list = (design_names,design_vals[1:100])
 

src/functions.jl

@@ -19,8 +19,8 @@ Computes a grid of prior probabilities over model parameters.
 - `parm_grid`: a grid of parameter values
 """
 function prior_probs(prior, parm_grid)
-    dens = [mapreduce((θ,d)->pdf(d, θ), *, g, prior) for g in parm_grid]
-    return dens/sum(dens)
+    dens = [mapreduce((θ,d) -> pdf(d, θ), *, g, prior) for g in parm_grid]
+    return dens / sum(dens)
 end
 
 """
@@ -32,7 +32,7 @@ Computes a grid of uniform prior probabilities over model parameters.
 - `parm_grid`: a grid of parameter values
 """
 function prior_probs(prior::Nothing, parm_grid)
-    return fill(1/length(parm_grid), size(parm_grid))
+    return fill(1 / length(parm_grid), size(parm_grid))
 end
 
 """
@@ -200,11 +200,11 @@ end
 # """
 # function marginal_posterior(optimizer)
 #     @unpack posteriors = optimizer
-#     return map(d->sum(posterior, dims=d), ndims(posterior):-1:1)
+#     return map(d-> sum(posterior, dims=d), ndims(posterior):-1:1)
 # end
 
 function compute_entropy(log_like)
-    return -1*sum(exp.(log_like) .* log_like, dims=3)[:,:]
+    return -1 * sum(exp.(log_like) .* log_like, dims=3)[:,:]
 end
 
 function conditional_entropy(entropy, post)
@@ -223,7 +223,7 @@ function marginal_entropy!(optimizer::Optimizer)
 end
 
 function marginal_entropy(marg_log_like)
-    return -sum(exp.(marg_log_like).*marg_log_like, dims=3)[:]
+    return -sum(exp.(marg_log_like) .* marg_log_like, dims=3)[:]
 end
 
 """
@@ -344,7 +344,7 @@ function to_grid(vals::NamedTuple)
 end
 
 function to_grid(vals)
-    k = [Symbol(string("v",i)) for i in 1:length(vals[1])]
+    k = [Symbol(string("v", i)) for i in 1:length(vals[1])]
     return (k...,),vals
 end
 
@@ -368,12 +368,12 @@ function mean_post(optimizer)
 end
 
 function mean_post(post, parm_grid)
-    return mapreduce((p,v)->p.*v, .+, post, parm_grid)
+    return mapreduce((p,v) -> p .* v, .+, post, parm_grid)
 end
 
 function std_post(post, parm_grid)
     mu = mean_post(post, parm_grid)
-    return mapreduce((p,v)->p.*(v .- mu).^2, .+, post, parm_grid) .|> sqrt
+    return mapreduce((p,v) -> p .* (v .- mu).^2, .+, post, parm_grid) .|> sqrt
 end
 
 function std_post(optimizer)
@@ -415,7 +415,7 @@ end
 # end
 
 # # function update_model_posterior(ps, es)
-# #     map(i->ps[i]./(sum(ps.*(es./es[i]))),1:length(ps))
+# #     map(i-> ps[i]./(sum(ps.*(es./es[i]))),1:length(ps))
 # # end
 
 # using Distributions

src/structs.jl

@@ -57,14 +57,16 @@ mutable struct Model{F<:Function,P}
 end
 
 function Model(args...; prior=nothing, loglike, kwargs...)
-    return Model(prior, (x...)->loglike(x..., args...; kwargs...))
+    return Model(prior, (x...) -> loglike(x..., args...; kwargs...))
 end
 
 """
     Optimizer(;task, model, grid_design, grid_parms, grid_response)
 
 `Optimizer` constructs a model object for adaptive design optimization
 
+# Fields 
+
 - `model`: a model object
 - `grid_design`:a grid of design parameters 
 - `grid_parms`: a grid of model parameters
@@ -74,7 +76,13 @@ end
 - `priors`: a multidimensional array of prior probabilities for parameters
 - `log_post`: a one dimensional array of log posterior probabilities for parameters
 - `entropy`: a two dimensional array of entropy values for parameter and design combinations
-- 
+- `marg_entropy`:
+- `cond_entropy`:
+- `mutual_info`:
+- `best_design`:
+- `parm_names`:
+- `model_state`:
+- `update_state!`:
 """
 mutable struct Optimizer{A,MT,M<:Model,T1,T2,T3,T4,T5,T6,T7,T8,T9,
     T10,T11}