experimenting with examples

Project.toml

@@ -23,6 +23,7 @@ OptimizationOptimJL = "36348300-93cb-4f02-beb5-3c3902f8871e"
 Plots = "91a5bcdd-55d7-5caf-9e0b-520d859cae80"
 Random = "9a3f8284-a2c9-5f02-9a11-845980a1fd5c"
 ReinforcementLearning = "158674fc-8238-5cab-b5ba-03dfc80d1318"
+ReinforcementLearningTrajectories = "6486599b-a3cd-4e92-a99a-2cea90cc8c3c"
 Statistics = "10745b16-79ce-11e8-11f9-7d13ad32a3b2"
 StatsPlots = "f3b207a7-027a-5e70-b257-86293d7955fd"
 

reinforecementLearning/K-armed Bandit.jl

@@ -1,4 +1,4 @@
-using ReinforcementLearning
+using ReinforcementLearning, ReinforcementLearningTrajectories
 using Plots
 
 # ╔═╡ 1fcd93f0-4a5c-11eb-252d-9da5bc78b08b
@@ -23,18 +23,19 @@ init=0.0
 opt= InvDecay(1.0)
 agent= Agent(
             policy=QBasedPolicy(
-                learner= MCLearner(
-                    approximator= TabularQApproximator(
+                learner= TabularQApproximator(
                         n_state= length(state_space(env)), 
                         n_action=length(action_space(env)), 
                         init= init, 
 
-                    ), 
-                    y=1.0
-                ),
+                    ),
                 explorer=explorer
             ),
             trajectory=VectorSARTTrajectory()
             )
             h1 = CollectBestAction(;best_Action=findmax(env.true_values)[2])
             h2 = TotalRewardPerEpisode(;is_display_on_exit=false)
+
+
+
+            t = Trajectory(Traces(a=Int[], b=Bool[]), BatchSampler(3), InsertSampleRatioControler(1.0, 3));

reinforecementLearning/One-dimensionalRandomWalk.jl

@@ -0,0 +1,17 @@
+using ReinforcementLearning
+
+
+env = RandomWalk1D()
+
+
+state_space(env)
+
+action_space(env)
+
+
+state(env)
+
+is_terminated(env)
+
+
+rand(2)