Add mutation to generate random tree

README.md

@@ -68,21 +68,28 @@ Larger alpha means more exploration.
 
 One can also adjust the relative probabilities of each operation here:
 ```julia
-weights = [8, 1, 1, 1, 0.1, 2]
+weights = [8, 1, 1, 1, 0.1, 0.5, 2]
 ```
-(for: 1. perturb constant, 2. mutate operator,
-3. append a node, 4. delete a subtree, 5. simplify equation,
-6. do nothing).
+for:
+
+1. Perturb constant
+2. Mutate operator
+3. Append a node
+4. Delete a subtree
+5. Simplify equation
+6. Randomize completely
+7. Do nothing
 
 
 # TODO
 
 - [ ] Hyperparameter tune
+- [ ] Add mutation for constant<->variable
 - [ ] Create a Python interface
 - [ ] Create a benchmark for accuracy
 - [ ] Create struct to pass through all hyperparameters, instead of treating as constants
     - Make sure doesn't affect performance
-- [ ] Use NN to generate weights over all probability distribution, and train on some randomly-generated equations
+- [ ] Use NN to generate weights over all probability distribution conditional on error and existing equation, and train on some randomly-generated equations
 - [ ] Performance:
     - [ ] Use an enum for functions instead of storing them?
     - Current most expensive operations:

eureqa.jl

@@ -334,7 +334,7 @@ function iterate(
 
     mutationChoice = rand()
     weight_for_constant = min(8, countConstants(tree))
-    weights = [weight_for_constant, 1, 1, 1, 0.1, 2] .* 1.0
+    weights = [weight_for_constant, 1, 1, 1, 0.1, 0.5, 2] .* 1.0
     weights /= sum(weights)
     cweights = cumsum(weights)
     n = countNodes(tree)
@@ -350,6 +350,8 @@ function iterate(
     elseif mutationChoice < cweights[5]
         tree = simplifyTree(tree) # Sometimes we simplify tree
         return tree
+    elseif mutationChoice < cweights[6]
+        tree = genRandomTree(5) # Sometimes we simplify tree
     else
         return tree
     end

hyperparams.jl

@@ -26,6 +26,8 @@ const migration = true
 const hofMigration = true
 # Fraction of population to replace with hall of fame
 const fractionReplacedHof = 0.1f0
+# Optimize constants
+const shouldOptimizeConstants = true
 ##################
 
 

paralleleureqa.jl

@@ -35,8 +35,10 @@ function fullRun(niterations::Integer;
         @inbounds Threads.@threads for i=1:nthreads
             allPops[i] = run(allPops[i], ncyclesperiteration, annealing, verbosity=verbosity)
             bestSubPops[i] = bestSubPop(allPops[i], topn=topn)
-            for j=1:bestSubPops[i].n
-                bestSubPops[i].members[j] = optimizeConstants(bestSubPops[i].members[j])
+            if shouldOptimizeConstants
+                for j=1:bestSubPops[i].n
+                    bestSubPops[i].members[j] = optimizeConstants(bestSubPops[i].members[j])
+                end
             end
         end
 
@@ -49,7 +51,6 @@ function fullRun(niterations::Integer;
             size = countNodes(member.tree)
             if member.score < hallOfFame.members[size].score
                 hallOfFame.members[size] = deepcopy(member)
-                #hallOfFame.members[size] = optimizeConstants(hallOfFame.members[size])
                 hallOfFame.exists[size] = true
             end
         end