update

src/interface.jl

@@ -9,7 +9,6 @@ using ..type, ..mcts, ..util
 # ---------------------------------------------- 100 --------------------------------------------- #
 
 
-
 """ Search the best action to take for a given state and task 
 
 # Arguments
@@ -63,14 +62,15 @@ function runMCTS(
   explorationweight::Number=1.0,
   earlystop::Union{Function,Nothing}=nothing,
   saveSimulatedNode::Bool=false,
-  multithread=false
-  )::NamedTuple{(:root, :bestNextState, :bestFinalState),Tuple{MCTSNode,T,T}} where {T<:Any}
+  multithread=false,
+  )::NamedTuple{(:root, :bestNextState, :bestTerminalState, :highValueStateList),
+                Tuple{MCTSNode,T,T,Vector{Any}}} where {T<:Any}
 
   root = MCTSNode("root", initialstate, 0, 0, 0, 0, false, nothing, Dict{String,MCTSNode}(),
                   Dict{Symbol,Any}())
 
-  # [WORKING] storage for holding all high reward terminal nodes
-  highStateValueNode = Channel{Any}(100)
+  # storage for holding all high reward terminal nodes
+  highValueState = Channel{Any}(100)
 
   for nth in 1:maxiterations
     node = root
@@ -99,7 +99,7 @@ function runMCTS(
                   horizontalSampleSimulationPhase=horizontalSampleSimulationPhase,
                   saveSimulatedNode=saveSimulatedNode,
                   multithread=multithread,
-                  highStateValueNode=highStateValueNode,
+                  highValueState=highValueState,
                   )
         end
       else
@@ -109,7 +109,7 @@ function runMCTS(
             horizontalSampleSimulationPhase=horizontalSampleSimulationPhase,
             saveSimulatedNode=saveSimulatedNode,
             multithread=multithread,
-            highStateValueNode=highStateValueNode)
+            highValueState=highValueState)
         end
       end
     end
@@ -120,15 +120,23 @@ function runMCTS(
     end
   end
 
-  # select the best next state and the best final state
+  # select the best next state and the best terminal state along the best trajectory
   bestNextState = selectBestNextNode(root)
-  besttrajectory = selectBestTrajectoryNode(root)
+  bestTerminalState = selectBestTrajectoryNode(root)
 
-  #[WORKING] compare all high value answer then select the best one
+  # take all high value state from highValueState channel and put it in a list
+  highValueStateList = Any[]
+  while !isempty(highValueState)
+    push!(highValueStateList, take!(highValueState))
+  end
 
-  return (root=root, bestNextState=bestNextState.state, bestFinalState=besttrajectory.state)
+  return (root=root, 
+          bestNextState=bestNextState.state, 
+          bestTerminalState=bestTerminalState.state,
+          highValueStateList=highValueStateList)
 end
 
+
 """ Search the best action to take for a given state and task 
 
 # Arguments
@@ -156,18 +164,18 @@ function simulateThenBackpropagate(node::MCTSNode, transition::Function, transit
                 maxSimulationDepth::Integer=3, horizontalSampleSimulationPhase::Integer=3,
                 saveSimulatedNode::Bool=false,
                 multithread=false,
-                highStateValueNode=Union{Nothing,Any}=nothing)
+                highValueState=Union{Nothing,Any}=nothing)
   simTrajectoryReward, terminalstate = 
     simulate(node, transition, transitionargs;
             maxSimulationDepth=maxSimulationDepth, 
             horizontalSample=horizontalSampleSimulationPhase,
             multithread=multithread)
-  # if a node has state value >= 8, store it in highStateValueNode
-  if highStateValueNode !== nothing && 
+  # if a node has state value >= 8, store it in highValueState
+  if highValueState !== nothing && 
     terminalstate !== nothing &&
     terminalstate[:reward] >= 8
 
-    put!(highStateValueNode, deepcopy(terminalstate))
+    put!(highValueState, deepcopy(terminalstate))
   end
 
   backpropagate(node, simTrajectoryReward)