-- LuigI/O by Akisame based on SethBling's MarI/O with modifications described here: https://pastebin.com/7qkeVVxi.
-- Pressing N will show the Neural net, pressing M will show the mutation rates, pressing L will load the last saved generation, pressing B will display the genome with the top fitness, T will activate the turbo and S will save the current location in your pool.
-- This is an early version and still a work in progress.
-- To use this script you will need to download FCEUX and create a savestate in state 1 as player 2(luigi) for smb. You will then need to go to config -> input and set player 2 to gamepad.
-- Click file -> lua -> 'new lua script window' and load the script.
-- Have fun!
-- Keep in mind that this code is provided for Non-Commercial usage and it is not permitted to request financial compensation (either by means of ads, donations, subscriptions or other means) for selling, streaming or otherwise previewing this code or its execution.
--FCEUX doesn't allow for user string input but if you press 'L' on your keyboard it should load the latest generation that was saved
--for loading specific generations replace nil with the filename as shown in the example below
savedpool = nil
--savedpool = "backup.71.SMB2-4.state.pool"
--savedpool = "backups/backup.118.SMB2-1.state.pool"
-- HUD options. These can be changed on the fly by pressing 'M' for the mutation rates and 'N' for the network
mutation_rates_disp = false
neural_net_disp = true
--set to false if you want warning popups before loading anything that might cause you to lose your current progress
nopopup = true
savestate_slot = 1 -- Load the level from FCEUX savestate slot.
--SAVE_LOAD_FILE = "test.pool" --filename to be used
SAVE_LOAD_FILE = nil --auto generate appropriate name from level values in ram (might not completely match with actual level)
player = 2 --1 for mario 2 for luigi
AllowAutoTurbo = true --Automatic turbo
AllowSlowdownNearMax = true --This is slow down the turbo when it nears to furthest it has currently gotten
FirstGenSpeedup = true --This will speed up the first generation
BottleneckTurbo = false --Activate turbo on bottleneck detection
TurboGenBottleneck = 20 --will start the turbo when it is in a bottleneck for 20 generations. Please adjust to match your taste
os.execute("mkdir backups") --create the folder to save the pools in
SavestateObj = savestate.object(savestate_slot)
ButtonNames = {
"A",
"B",
"up",
"down",
"left",
"right",
}
BoxRadius = 6
InputSize = (BoxRadius*2+1)*(BoxRadius*2+1)
switchtime = {10,12,18}
switchtimer = {}
initialswitchvalue = {}
for i=1,#switchtime do
switchtimer[i] = switchtime[i]
initialswitchvalue[i] = 1
end
Inputs = InputSize+1+#initialswitchvalue
Outputs = #ButtonNames
Population = 100
DeltaDisjoint = 2.0
DeltaWeights = 0.4
DeltaThreshold = 1.0
keyflag = false --used for keyboard input
endlevel = false
preparenext = false
secondbest = 0
maxright = 0
restoreturbo = false
rerunning = false
killcounter = 0
StaleSpecies = 25 --increased slightly to allow for longer evolution
mariohole = false --flag for mario falling into a hole
mariopipe = false --flag for mario exiting a pipe
marioPipeEnter = false --new flag for entering a pipe
offset = 0 --offset value for when mario exits a pipe
timebonus = 0 --timebonus to prevent the hard drops in fitness when exiting a pipe
maxcounter = 0 --species that reach the max fitness
currentTracker = 0 --tracks location on screen
loop = false --flag for when a loop is detected
previousmaxfitness = 0
bottleneckcounter = 0
MutateConnectionsChance = 0.25
PerturbChance = 0.90
CrossoverChance = 0.75
LinkMutationChance = 2.0
NodeMutationChance = 0.50
BiasMutationChance = 0.40
StepSize = 0.1
DisableMutationChance = 0.4
EnableMutationChance = 0.2
oscilationmutationchance = 0.2
interbreedchance = 0.03
networkswitchmutationchance = 0.01
replaceSecondNetworkChance = 0.01
nswitch = true
ntrigger = true
ncount = 0
TimeoutConstant = 120 --set to a higher value but see for yourself what is acceptable
MaxNodes = 1000000
function getPositions()
marioX = memory.readbyte(0x6D) * 0x100 + memory.readbyte(0x86)
marioY = memory.readbyte(0x03B8)+16
mariostate = memory.readbyte(0x0E) --get mario's state. (entering exiting pipes, dying, picking up a mushroom etc etc)
yspeed = memory.readbyte(0x009F) --y velocity 1-5 are falling and 250-255 is jumping
if yspeed >1 and yspeed <10 then
falling = true
else
falling = false
end
currentscreen = memory.readbyte(0x071A) --finds current screen for loop detection
nextscreen = memory.readbyte(0x071B) --finds next screen
CurrentWorld = memory.readbyte(0x075F) --finds the current world (value + 1 is world)
CurrentLevel = memory.readbyte(0x0760) --finds the current level (0=1 2=2 3=3 4=4)
demoruncheck = memory.readbyte(0x0770) --equals 0 for demo and 1 for normal run
statuscheck = memory.readbyte(0x0772) -- 3= playing 1 = loading
screenX = memory.readbyte(0x03AD)
screenY = memory.readbyte(0x03B8)
end
function getTile(dx, dy)
local x = marioX + dx + 8
local y = marioY + dy - 16
local page = math.floor(x/256)%2
local subx = math.floor((x%256)/16)
local suby = math.floor((y - 32)/16)
local addr = 0x500 + page*13*16+suby*16+subx
if suby >= 13 or suby < 0 then
return 0
end
if memory.readbyte(addr) ~= 0 then
return 1
else
return 0
end
end
function getSprites()
local sprites = {}
for slot=0,4 do
local enemy = memory.readbyte(0xF+slot)
if enemy ~= 0 then
local ex = memory.readbyte(0x6E + slot)*0x100 + memory.readbyte(0x87+slot)
local ey = memory.readbyte(0xCF + slot)+36 --changed this to stop sprites from floating in mid air on the neural network
sprites[#sprites+1] = {["x"]=ex,["y"]=ey}
end
end
return sprites
end
function getInputs()
getPositions()
sprites = getSprites()
local inputs = {}
for dy=-BoxRadius*16,BoxRadius*16,16 do
for dx=-BoxRadius*16,BoxRadius*16,16 do
inputs[#inputs+1] = 0
tile = getTile(dx, dy)
if tile == 1 and marioY+dy < 0x1B0 then
inputs[#inputs] = 1
end
for i = 1,#sprites do
distx = math.abs(sprites[i]["x"] - (marioX+dx))
disty = math.abs(sprites[i]["y"] - (marioY+dy+16))
if distx <= 8 and disty <= 8 then
inputs[#inputs] = -1
end
end
end
end
return inputs
end
function sigmoid(x)
return 2/(1+math.exp(-4.9*x))-1
end
function newInnovation()
pool.innovation = pool.innovation + 1
return pool.innovation
end
function newPool()
local pool = {}
pool.species = {}
pool.generation = 0
pool.innovation = Outputs
pool.currentSpecies = 1
pool.currentGenome = 1
pool.currentFrame = 0
pool.maxFitness = 0
pool.maxcounter = 0 --counter for %max species reached
pool.averagemaxfitness = 0
return pool
end
function newSpecies()
local species = {}
species.topFitness = 0
species.staleness = 0
species.genomes = {}
species.averageFitness = 0
species.nickname = 'none'
species.turbo = 'on'
return species
end
function getCurrentGenome()
local species = pool.species[pool.currentSpecies]
return species.genomes[pool.currentGenome]
end
function toRGBA(ARGB)
return bit.lshift(ARGB, 8) + bit.rshift(ARGB, 24)
end
function newGenome()
local genome = {}
genome.genes = {}
genome.genes2 = {}
genome.fitness = 0
genome.adjustedFitness = 0
genome.network = {}
genome.network2 = {}
genome.maxneuron = 0
genome.globalRank = 0
genome.oscilations = {}
genome.networkswitch = 0
for i=1,#switchtime do
genome.oscilations[i] = switchtime[i]
end
genome.mutationRates = {}
genome.mutationRates["connections"] = MutateConnectionsChance
genome.mutationRates["link"] = LinkMutationChance
genome.mutationRates["bias"] = BiasMutationChance
genome.mutationRates["node"] = NodeMutationChance
genome.mutationRates["enable"] = EnableMutationChance
genome.mutationRates["disable"] = DisableMutationChance
genome.mutationRates["step"] = StepSize
genome.mutationRates["oscilation"] = oscilationmutationchance
genome.mutationRates["networkswitch"] = networkswitchmutationchance
return genome
end
function copyGenome(genome)
local genome2 = newGenome()
for g=1,#genome.genes do
table.insert(genome2.genes, copyGene(genome.genes[g]))
end
for g=1,#genome.genes2 do
table.insert(genome2.genes2, copyGene(genome.genes2[g]))
end
for i=1,#genome.oscilations do
genome2.oscilations[i] = genome.oscilations[i]
end
genome2.networkswitch = genome.networkswitch
genome2.maxneuron = genome.maxneuron
genome2.mutationRates["connections"] = genome.mutationRates["connections"]
genome2.mutationRates["link"] = genome.mutationRates["link"]
genome2.mutationRates["bias"] = genome.mutationRates["bias"]
genome2.mutationRates["node"] = genome.mutationRates["node"]
genome2.mutationRates["enable"] = genome.mutationRates["enable"]
genome2.mutationRates["disable"] = genome.mutationRates["disable"]
genome2.mutationRates["oscilation"] = genome.mutationRates["oscilation"]
genome2.mutationRates["networkswitch"] = genome.mutationRates["networkswitch"]
return genome2
end
function basicGenome()
local genome = newGenome()
local innovation = 1
genome.maxneuron = Inputs
mutate(genome, 1)
mutate(genome, 2)
return genome
end
function newGene()
local gene = {}
gene.into = 0
gene.out = 0
gene.weight = 0.0
gene.enabled = true
gene.innovation = 0
return gene
end
function copyGene(gene)
local gene2 = newGene()
gene2.into = gene.into
gene2.out = gene.out
gene2.weight = gene.weight
gene2.enabled = gene.enabled
gene2.innovation = gene.innovation
return gene2
end
function newNeuron()
local neuron = {}
neuron.incoming = {}
neuron.value = 0.0
neuron.switcher = false
neuron.timer = 60
neuron.multiplier = 1
return neuron
end
function generateNetwork(genome)
local network = {}
network.neurons = {}
for i=1,Inputs do
network.neurons[i] = newNeuron()
end
for o=1,Outputs do
network.neurons[MaxNodes+o] = newNeuron()
end
table.sort(genome.genes, function (a,b)
return (a.out < b.out)
end)
for i=1,#genome.genes do
local gene = genome.genes[i]
if gene.enabled then
if network.neurons[gene.out] == nil then
network.neurons[gene.out] = newNeuron()
end
local neuron = network.neurons[gene.out]
table.insert(neuron.incoming, gene)
if network.neurons[gene.into] == nil then
network.neurons[gene.into] = newNeuron()
end
end
end
genome.network = network
local network2 = {}
network2.neurons = {}
for i=1,Inputs do
network2.neurons[i] = newNeuron()
end
for o=1,Outputs do
network2.neurons[MaxNodes+o] = newNeuron()
end
table.sort(genome.genes2, function (a,b)
return (a.out < b.out)
end)
for i=1,#genome.genes2 do
local gene = genome.genes2[i]
if gene.enabled then
if network2.neurons[gene.out] == nil then
network2.neurons[gene.out] = newNeuron()
end
local neuron = network2.neurons[gene.out]
table.insert(neuron.incoming, gene)
if network2.neurons[gene.into] == nil then
network2.neurons[gene.into] = newNeuron()
end
end
end
genome.network2 = network2
end
function evaluateNetwork(network, inputs, genome)
table.insert(inputs, 1) -- bias input node
for i = 1,#initialswitchvalue do
if genome.oscilations[i] > 0 then
if switchtimer[i] > 0 then
switchtimer[i] = switchtimer[i] -1
else
switchtimer[i] = genome.oscilations[i]
initialswitchvalue[i] = initialswitchvalue[i]*-1
end
end
table.insert(inputs, initialswitchvalue[i])
end
if #inputs ~= Inputs then
emu.print("Incorrect number of neural network inputs.")
return {}
end
for i=1,Inputs do
network.neurons[i].value = inputs[i]
end
for _,neuron in pairs(network.neurons) do
local sum = 0
for j = 1,#neuron.incoming do
local incoming = neuron.incoming[j]
local other = network.neurons[incoming.into]
sum = sum + incoming.weight * other.value
end
if #neuron.incoming > 0 then
if neuron.switcher and sigmoid(sum)>0 then
neuron.timer = neuron.timer - 1
elseif neuron.switcher then
neuron.timer = 120
neuron.multiplier = 1
end
if neuron.switcher and neuron.timer < 0 then
neuron.multiplier = -neuron.multiplier
neuron.timer = 10
end
neuron.value = sigmoid(sum) * neuron.multiplier
end
end
local outputs = {}
for o=1,Outputs do
local button = ButtonNames[o]
if network.neurons[MaxNodes+o].value > 0 then
outputs[button] = true
else
outputs[button] = false
end
end
return outputs
end
function crossover(g1, g2)
-- Make sure g1 is the higher fitness genome
if g2.fitness > g1.fitness then
tempg = g1
g1 = g2
g2 = tempg
end
local child = newGenome()
for i=1,#g1.oscilations do
child.oscilations[i] = g1.oscilations[i]
end
child.networkswitch = g1.networkswitch
local innovations2 = {}
for i=1,#g2.genes do
local gene = g2.genes[i]
innovations2[gene.innovation] = gene
end
for i=1,#g1.genes do
local gene1 = g1.genes[i]
local gene2 = innovations2[gene1.innovation]
if gene2 ~= nil and math.random(1,2) == 1 and gene2.enabled then
table.insert(child.genes, copyGene(gene2))
else
table.insert(child.genes, copyGene(gene1))
end
end
for i=1,#g1.genes2 do
local gene1 = g1.genes2[i]
local gene2 = innovations2[gene1.innovation]
if gene2 ~= nil and math.random(1,2) == 1 and gene2.enabled then
table.insert(child.genes2, copyGene(gene2))
else
table.insert(child.genes2, copyGene(gene1))
end
end
child.maxneuron = math.max(g1.maxneuron,g2.maxneuron)
for mutation,rate in pairs(g1.mutationRates) do
child.mutationRates[mutation] = rate
end
return child
end
function randomNeuron(genes, nonInput)
local neurons = {}
if not nonInput then
for i=1,Inputs do
neurons[i] = true
end
end
for o=1,Outputs do
neurons[MaxNodes+o] = true
end
for i=1,#genes do
if (not nonInput) or genes[i].into > Inputs then
neurons[genes[i].into] = true
end
if (not nonInput) or genes[i].out > Inputs then
neurons[genes[i].out] = true
end
end
local count = 0
for _,_ in pairs(neurons) do
count = count + 1
end
local n = math.random(1, count)
for k,v in pairs(neurons) do
n = n-1
if n == 0 then
return k
end
end
return 0
end
function containsLink(genes, link)
for i=1,#genes do
local gene = genes[i]
if gene.into == link.into and gene.out == link.out then
return true
end
end
end
function pointMutate(genome, which)
local step = genome.mutationRates["step"]
local gene = {}
if which == 1 then
for i=1,#genome.genes do
gene = genome.genes[i]
if math.random() < PerturbChance then
gene.weight = gene.weight + math.random() * step*2 - step
else
gene.weight = math.random()*4-2
end
end
else
for i=1,#genome.genes2 do
gene = genome.genes2[i]
if math.random() < PerturbChance then
gene.weight = gene.weight + math.random() * step*2 - step
else
gene.weight = math.random()*4-2
end
end
end
end
function linkMutate(genome, forceBias, which)
local neuron1 = 0
local neuron2 = 0
if which == 1 then
neuron1 = randomNeuron(genome.genes, false)
neuron2 = randomNeuron(genome.genes, true)
else
neuron1 = randomNeuron(genome.genes2, false)
neuron2 = randomNeuron(genome.genes2, true)
end
local newLink = newGene()
if neuron1 <= Inputs and neuron2 <= Inputs then
--Both input nodes
return
end
if neuron2 <= Inputs then
-- Swap output and input
local temp = neuron1
neuron1 = neuron2
neuron2 = temp
end
newLink.into = neuron1
newLink.out = neuron2
if forceBias then
local adjustment = math.random(0,#initialswitchvalue)
newLink.into = Inputs - adjustment
end
if which == 1 then
if containsLink(genome.genes, newLink) then
return
end
else
if containsLink(genome.genes2, newLink) then
return
end
end
newLink.innovation = newInnovation()
newLink.weight = math.random()*4-2
if which == 1 then
table.insert(genome.genes, newLink)
else
table.insert(genome.genes2, newLink)
end
end
function nodeMutate(genome, which)
if which == 1 then
if #genome.genes == 0 then
return
end
else
if #genome.genes2 == 0 then
return
end
end
genome.maxneuron = genome.maxneuron + 1
local gene = {}
if which == 1 then
gene = genome.genes[math.random(1,#genome.genes)]
else
gene = genome.genes2[math.random(1,#genome.genes2)]
end
if not gene.enabled then
return
end
gene.enabled = false
local gene1 = copyGene(gene)
gene1.out = genome.maxneuron
gene1.weight = 1.0
gene1.innovation = newInnovation()
gene1.enabled = true
if which == 1 then
table.insert(genome.genes, gene1)
else
table.insert(genome.genes2, gene1)
end
local gene2 = copyGene(gene)
gene2.into = genome.maxneuron
gene2.innovation = newInnovation()
gene2.enabled = true
if which == 1 then
table.insert(genome.genes, gene2)
else
table.insert(genome.genes2, gene2)
end
end
function enableDisableMutate(genome, enable, which)
local candidates = {}
if which == 1 then
for _,gene in pairs(genome.genes) do
if gene.enabled == not enable then
table.insert(candidates, gene)
end
end
else
for _,gene in pairs(genome.genes2) do
if gene.enabled == not enable then
table.insert(candidates, gene)
end
end
end
if #candidates == 0 then
return
end
local gene = candidates[math.random(1,#candidates)]
gene.enabled = not gene.enabled
end
function oscilationMutate(genome)
mutationpoint = math.random(1,#genome.oscilations)
if genome.oscilations[mutationpoint]>0 then
if math.random(1,2) == 1 then
genome.oscilations[mutationpoint] = genome.oscilations[mutationpoint] + 1
else
genome.oscilations[mutationpoint] = genome.oscilations[mutationpoint] - 1
end
end
if genome.oscilations[mutationpoint]<0 then
genome.oscilations[mutationpoint] = 0
end
end
function NswitchMutate(genome)
if math.random(1,2) == 1 then
if math.random(1,2) == 1 then
genome.networkswitch = genome.networkswitch + 0.1
else
genome.networkswitch = genome.networkswitch - 0.1
end
else
if math.random(1,2) == 1 then
genome.networkswitch = genome.networkswitch + 0.5
else
genome.networkswitch = genome.networkswitch - 0.5
end
end
if genome.networkswitch < 1 then
genome.networkswitch = math.random(0,8)
end
if genome.networkswitch > 8 then
genome.networkswitch = math.random(0,8)
end
end
function mutate(genome, which)
for mutation,rate in pairs(genome.mutationRates) do
if math.random(1,2) == 1 then
genome.mutationRates[mutation] = 0.95*rate
else
genome.mutationRates[mutation] = 1.05263*rate
end
end
if math.random() < genome.mutationRates["connections"] then
pointMutate(genome, which)
end
local p = genome.mutationRates["link"]
while p > 0 do
if math.random() < p then
linkMutate(genome, false, which)
end
p = p - 1
end
p = genome.mutationRates["bias"]
while p > 0 do
if math.random() < p then
linkMutate(genome, true, which)
end
p = p - 1
end
p = genome.mutationRates["node"]
while p > 0 do
if math.random() < p then
nodeMutate(genome, which)
end
p = p - 1
end
p = genome.mutationRates["enable"]
while p > 0 do
if math.random() < p then
enableDisableMutate(genome, true, which)
end
p = p - 1
end
p = genome.mutationRates["disable"]
while p > 0 do
if math.random() < p then
enableDisableMutate(genome, false, which)
end
p = p - 1
end
p = genome.mutationRates["oscilation"]
while p > 0 do
if math.random() < p then
oscilationMutate(genome)
end
p = p - 1
end
if genome.networkswitch > 0 then
p = genome.mutationRates["networkswitch"]
while p > 0 do
if math.random() < p then
NswitchMutate(genome)
end
p = p - 1
end
end
end
function disjoint(genes1, genes2)
local i1 = {}
for i = 1,#genes1 do
local gene = genes1[i]
i1[gene.innovation] = true
end
local i2 = {}
for i = 1,#genes2 do
local gene = genes2[i]
i2[gene.innovation] = true
end
local disjointGenes = 0
for i = 1,#genes1 do
local gene = genes1[i]
if not i2[gene.innovation] then
disjointGenes = disjointGenes+1
end
end
for i = 1,#genes2 do
local gene = genes2[i]
if not i1[gene.innovation] then
disjointGenes = disjointGenes+1
end
end
local n = math.max(#genes1, #genes2)
return disjointGenes / n
end
function weights(genes1, genes2)
local i2 = {}
for i = 1,#genes2 do
local gene = genes2[i]
i2[gene.innovation] = gene
end
local sum = 0
local coincident = 0
for i = 1,#genes1 do
local gene = genes1[i]
if i2[gene.innovation] ~= nil then
local gene2 = i2[gene.innovation]
sum = sum + math.abs(gene.weight - gene2.weight)
coincident = coincident + 1
end
end
return sum / coincident
end
function sameSpecies(genome1, genome2)
if genome1.networkswitch == 0 and genome2.networkswitch == 0 then
local dd = DeltaDisjoint*disjoint(genome1.genes, genome2.genes)
local dw = DeltaWeights*weights(genome1.genes, genome2.genes)
return dd + dw < DeltaThreshold
elseif math.abs(genome1.networkswitch - genome2.networkswitch) > 0.6 then
return false
else
local dd = DeltaDisjoint*disjoint(genome1.genes, genome2.genes)
local dw = DeltaWeights*weights(genome1.genes, genome2.genes)
local dd2 = DeltaDisjoint*disjoint(genome1.genes2, genome2.genes2)
local dw2 = DeltaWeights*weights(genome1.genes2, genome2.genes2)
if dd + dw < DeltaThreshold and dd2 + dw2 < DeltaThreshold then
return true
else
return false
end
end
end
function rankGlobally()
local global = {}
for s = 1,#pool.species do
local species = pool.species[s]
for g = 1,#species.genomes do
table.insert(global, species.genomes[g])
end
end
table.sort(global, function (a,b)
return (a.fitness < b.fitness)
end)
for g=1,#global do
global[g].globalRank = g
end
end
function calculateAverageFitness(species)
local total = 0
for g=1,#species.genomes do
local genome = species.genomes[g]
total = total + genome.globalRank
end
if species.topFitness < pool.averagemaxfitness - (pool.standardDeviationMaxFitness *2) then
species.averageFitness = (total / #species.genomes) / 4
elseif species.topFitness < pool.averagemaxfitness - pool.standardDeviationMaxFitness then
species.averageFitness = (total / #species.genomes) / 2.5
elseif species.topFitness > pool.averagemaxfitness + (pool.standardDeviationMaxFitness *3) then
species.averageFitness = (total / #species.genomes) * 2
elseif species.topFitness > pool.averagemaxfitness + (pool.standardDeviationMaxFitness *2) then
species.averageFitness = (total / #species.genomes) * 1.5
else
species.averageFitness = total / #species.genomes
end
end
function calculateStandardDeviationMax()
local deviationtotal = 0
for s=1,#pool.species do
deviationtotal = deviationtotal + math.abs(pool.species[s].topFitness - pool.averagemaxfitness)
end
pool.standardDeviationMaxFitness = deviationtotal / #pool.species
end
function AverageMaxFitness()
local total = 0
for s = 1,#pool.species do
local species = pool.species[s]
total = total + species.topFitness
end
pool.averagemaxfitness = total / #pool.species
end
function totalAverageFitness()
local total = 0
for s = 1,#pool.species do
local species = pool.species[s]
total = total + species.averageFitness
end
return total
end
function cullSpecies(cutToOne)
for s = 1,#pool.species do
local species = pool.species[s]
table.sort(species.genomes, function (a,b)
return (a.fitness > b.fitness)
end)
local remaining = math.ceil(#species.genomes/2)
if cutToOne then
remaining = 1
end
while #species.genomes > remaining do
table.remove(species.genomes)
end
end
end
function breedChild(species)
local child = {}
if math.random() < CrossoverChance then
g1 = species.genomes[math.random(1, #species.genomes)]
g2 = species.genomes[math.random(1, #species.genomes)]
child = crossover(g1, g2)
else
g = species.genomes[math.random(1, #species.genomes)]
child = copyGenome(g)
end
if child.networkswitch > 0 then
if math.random(1,2) == 1 then
mutate(child, 1)
else
mutate(child, 2)
end
else
mutate(child, 1)
end
return child
end
function secondbestspecies() --added this to help prevent the formation of monospecies when LuigI/O encounters a major bottleneck
local bestfitness = 0
local secondbestfitness = 0
for s = 1,#pool.species do
local species = pool.species[s]
if species.topFitness > bestfitness then
secondbestfitness = bestfitness
bestfitness = species.topFitness
elseif species.topFitness > secondbestfitness then
secondbestfitness = species.topFitness
end
end
return secondbestfitness
end
function removeStaleSpecies()
local survived = {}
secondbest = secondbestspecies()
emu.print("Max Fitness: ".. pool.maxFitness ..". second best: ".. secondbest)
for s = 1,#pool.species do
local species = pool.species[s]
table.sort(species.genomes, function (a,b)
return (a.fitness > b.fitness)
end)
if species.genomes[1].fitness > species.topFitness then
species.topFitness = species.genomes[1].fitness
species.staleness = 0
else
species.staleness = species.staleness + 1
end
if species.staleness < StaleSpecies or species.topFitness >= secondbest then --originally species.topFitness >= pool.maxFitness
table.insert(survived, species)
end
end
pool.species = survived
end
function removeWeakSpecies()
local survived = {}
local sum = totalAverageFitness()
for s = 1,#pool.species do
local species = pool.species[s]
breed = math.floor(species.averageFitness / sum * Population)
if breed >= 1 then
table.insert(survived, species)
end
end
pool.species = survived
end
function addToSpecies(child)
local foundSpecies = false
for s=1,#pool.species do
local species = pool.species[s]
if not foundSpecies and sameSpecies(child, species.genomes[1]) then
table.insert(species.genomes, child)
foundSpecies = true
end
end
if not foundSpecies then
local childSpecies = newSpecies()
table.insert(childSpecies.genomes, child)
table.insert(pool.species, childSpecies)
end
end
function replaceSecondNetwork(g1,g2)
local genome2 = newGenome()
for g=1,#g1.genes do
table.insert(genome2.genes, copyGene(g1.genes[g]))
end
for i=1,#g1.oscilations do
genome2.oscilations[i] = g1.oscilations[i]
end
genome2.networkswitch = g1.networkswitch
genome2.maxneuron = g1.maxneuron
genome2.mutationRates["connections"] = g1.mutationRates["connections"]
genome2.mutationRates["link"] = g1.mutationRates["link"]
genome2.mutationRates["bias"] = g1.mutationRates["bias"]
genome2.mutationRates["node"] = g1.mutationRates["node"]
genome2.mutationRates["enable"] = g1.mutationRates["enable"]
genome2.mutationRates["disable"] = g1.mutationRates["disable"]
genome2.mutationRates["oscilation"] = g1.mutationRates["oscilation"]
genome2.mutationRates["networkswitch"] = g1.mutationRates["networkswitch"]
if g2.networkswitch > 0 then
if math.random(1,2) == 1 then
for g=1,#g2.genes do
table.insert(genome2.genes2, copyGene(g2.genes[g]))
end
else
for g=1,#g2.genes2 do
table.insert(genome2.genes2, copyGene(g2.genes2[g]))
end
end
else
for g=1,#g2.genes do
table.insert(genome2.genes2, copyGene(g2.genes[g]))
end
end
if genome2.networkswitch == 0 then
local switch = math.pow(math.floor(g1.fitness/500),2)
genome2.networkswitch = math.ceil(math.sqrt(math.random(1,switch)))+(math.random(0,9)/10)
end
print('network switch: '.. genome2.networkswitch)
print('end number of genes2: '.. #genome2.genes2)
return genome2
end
function newGeneration()
if pool.maxFitness > previousmaxfitness then
previousmaxfitness = pool.maxFitness
bottleneckcounter = 0
if AllowAutoTurbo then
emu.speedmode("normal")
turbo = false
end
else
bottleneckcounter = bottleneckcounter + 1
end
if bottleneckcounter < 7 and Population > 300 then
if pool.generation < 5 then
Population = math.floor(Population * 0.9)
else
Population = math.floor(Population * 0.95)
end
if Population < 300 then
Population = 300
end
elseif bottleneckcounter > 10 and Population <900 then
Population = math.floor(Population * 1.10)
if Population > 900 then
Population = 900
end
end
if bottleneckcounter > 25 then
bottleneckcounter = 0
end
if bottleneckcounter > TurboGenBottleneck and AllowAutoTurbo and BottleneckTurbo then
emu.speedmode("turbo")
turbo = true
end
cullSpecies(false) -- Cull the bottom half of each species
rankGlobally()
removeStaleSpecies()
AverageMaxFitness()
calculateStandardDeviationMax()
rankGlobally()
for s = 1,#pool.species do
local species = pool.species[s]
calculateAverageFitness(species)
end
removeWeakSpecies()
local sum = totalAverageFitness()
local children = {}
for s = 1,#pool.species do
local species = pool.species[s]
breed = math.floor(species.averageFitness / sum * Population) - 1
if breed > 10 then
breed = math.ceil(breed * math.sqrt(species.topFitness / pool.maxFitness))
end
if breed > Population/4 then
breed = Population/4
end
for i=1,breed do
table.insert(children, breedChild(species))
end
end
cullSpecies(true) -- Cull all but the top member of each species
if #pool.species < 30 then
interbreedchance = 0.2
else
interbreedchance = 0.03
end
for i=1,#pool.species do
if math.random() < interbreedchance then
local counterbreed = math.random(1, #pool.species)
if counterbreed ~= i then
table.insert(children, crossover(pool.species[i].genomes[1], pool.species[counterbreed].genomes[1]))
end
end
if bottleneckcounter > 23 then
if math.random() < 0.1 then
local tfit = pool.species[i].topFitness
child = copyGenome(pool.species[i].genomes[1])
switchlocation = math.floor(tfit/500)
if child.networkswitch == 0 then
if math.random(1,2) == 1 then
switchlocation = switchlocation + (math.random(0,10)/10)
end
child.networkswitch = switchlocation
table.insert(children, child)
end
end
end
if math.random() < replaceSecondNetworkChance and pool.species[i].topFitness > 500 then
if pool.species[i].genomes[1].networkswitch > 0 then
if math.random() < 0.5 then
local donornetwork = math.random(1, #pool.species)
if donornetwork ~= i then
table.insert(children, replaceSecondNetwork(pool.species[i].genomes[1], pool.species[donornetwork].genomes[1]))
end
end
else
local donornetwork = math.random(1, #pool.species)
if donornetwork ~= i then
table.insert(children, replaceSecondNetwork(pool.species[i].genomes[1], pool.species[donornetwork].genomes[1]))
end
end
end
end
while #children + #pool.species < Population do
local species = pool.species[math.random(1, #pool.species)]
table.insert(children, breedChild(species))
end
for c=1,#children do
local child = children[c]
addToSpecies(child)
end
local children = {}
local rem_gen = 0
for s=1,#pool.species do
local species = pool.species[s]
if species.topFitness < secondbest and #species.genomes > Population/12 then
while #species.genomes > Population/12 do
table.remove(species.genomes)
rem_gen = rem_gen +1
end
elseif species.topFitness >= secondbest and #species.genomes>Population/7 then
while #species.genomes > Population/7 do
table.remove(species.genomes)
rem_gen = rem_gen +1
end
end
end
for x=1,rem_gen do
local species = pool.species[math.random(1, #pool.species)]
table.insert(children, breedChild(species))
end
for c=1,#children do
local child = children[c]
addToSpecies(child)
end
for s=1,#pool.species do
local species = pool.species[s]
if species.topFitness < secondbest and #species.genomes > Population/12 then
while #species.genomes > Population/12 do
table.remove(species.genomes)
end
elseif species.topFitness >= secondbest and #species.genomes>Population/7 then
while #species.genomes > Population/7 do
table.remove(species.genomes)
end
end
end
pool.generation = pool.generation + 1
pool.maxcounter = 0 --reset max fitness counter
writeFile("backups/backup." .. pool.generation .. "." .. SAVE_LOAD_FILE)
writelatestgen() --tracker for the latest generation
end
function initializePool()
pool = newPool()
for i=1,Population do
basic = basicGenome()
addToSpecies(basic)
end
initializeRun()
end
function clearJoypad()
controller = {}
for b = 1,#ButtonNames do
controller[ButtonNames[b]] = false
end
joypad.set(player, controller)
end
function initializeRun()
savestate.load(SavestateObj);
rightmost = 0
pool.currentFrame = 0
timeout = TimeoutConstant
clearJoypad()
currentTracker = memory.readbyte(0x071A) --sets the current tracker to the current screen
local species = pool.species[pool.currentSpecies]
local genome = species.genomes[pool.currentGenome]
for i=1,#genome.oscilations do
switchtimer[i] = genome.oscilations[i]
initialswitchvalue[i] = 1
end
generateNetwork(genome)
evaluateCurrent()
end
function reRun()
savestate.load(savestate.object(previous_savestate));
rightmost = 0
pool.currentFrame = 0
timeout = TimeoutConstant
clearJoypad()
currentTracker = memory.readbyte(0x071A) --sets the current tracker to the current screen
offset = 0 --offset in x coordinates for pipe
timebonus = 0 --used to freeze fitness
mariohole = false --reset the fallen into hole variable
mariopipe = false --is mario exiting a pipe or just teleporting
loop = false --is this a loop?
marioPipeEnter = false --has mario entered a pipe
currentTracker = 1 --tracker for the loop
killtrigger = false
ntrigger = true
nswitch = true
local species = pool.species[pool.currentSpecies]
local genome = species.genomes[pool.currentGenome]
for i=1,#genome.oscilations do
switchtimer[i] = genome.oscilations[i]
initialswitchvalue[i] = 1
end
generateNetwork(genome)
evaluateCurrent()
end
function evaluateCurrent()
local species = pool.species[pool.currentSpecies]
local genome = species.genomes[pool.currentGenome]
inputs = getInputs()
if nswitch then
controller = evaluateNetwork(genome.network, inputs, genome)
else
controller = evaluateNetwork(genome.network2, inputs, genome)
end
if controller["left"] and controller["right"] then
controller["left"] = false
controller["right"] = false
end
if controller["up"] and controller["down"] then
controller["up"] = false
controller["down"] = false
end
joypad.set(player, controller)
end
if pool == nil then
initializePool() --------------------------------------------------------------------------------------------------------------------------------------------------------
end
function nextGenome()
pool.currentGenome = pool.currentGenome + 1
if pool.currentGenome > #pool.species[pool.currentSpecies].genomes then
pool.currentGenome = 1
pool.currentSpecies = pool.currentSpecies+1
if pool.currentSpecies > #pool.species then
newGeneration()
pool.currentSpecies = 1
end
end
end
function fitnessAlreadyMeasured()
local species = pool.species[pool.currentSpecies]
local genome = species.genomes[pool.currentGenome]
return genome.fitness ~= 0
end
function GenerateNewKeepBest(genome)
Population = 1000
previousbest = copyGenome(genome)
previousmaxfitness = 1
maxright = 1
offset = 0 --offset in x coordinates for pipe
timebonus = 0 --used to freeze fitness
mariohole = false --reset the fallen into hole variable
mariopipe = false --is mario exiting a pipe or just teleporting
loop = false --is this a loop?
marioPipeEnter = false --has mario entered a pipe
currentTracker = 1 --tracker for the loop
pool = newPool()
addToSpecies(previousbest)
for i=1,Population-1 do
basic = basicGenome()
addToSpecies(basic)
end
initializeRun()
end
function displayGenome(genome)
gui.opacity(0.53)
if not genome then return end
local network = {}
if nswitch then
network = genome.network
else
network = genome.network2
end
local cells = {}
local i = 1
local cell = {}
for dy=-BoxRadius,BoxRadius do
for dx=-BoxRadius,BoxRadius do
cell = {}
cell.x = 50+5*dx
cell.y = 70+5*dy
cell.value = network.neurons[i].value
cells[i] = cell
i = i + 1
end
end
local biasCell = {}
biasCell.x = 80
biasCell.y = 110
biasCell.value = network.neurons[Inputs-#initialswitchvalue].value
cells[Inputs-#initialswitchvalue] = biasCell
for i=0,#initialswitchvalue-1 do
local switchCell = {}
switchCell.x = 80-10*#initialswitchvalue+10*i
switchCell.y = 110
switchCell.value = network.neurons[Inputs-i].value
cells[Inputs-i] = switchCell
end
gui.drawbox(215,32,245,82,toRGBA(0x80808080),toRGBA(0x00000000))
gui.opacity(0.85)
for o = 1,Outputs do
cell = {}
cell.x = 220
cell.y = 30 + 8 * o
cell.value = network.neurons[MaxNodes + o].value
cells[MaxNodes+o] = cell
local color
if cell.value > 0 then
color = 0xFF0000FF
else
color = 0xFF777777-- original color = 0xFF000000
end
gui.drawtext(225, 26+8*o, ButtonNames[o], toRGBA(color), 0x0) --moved slightly for better alignment
end
for n,neuron in pairs(network.neurons) do
cell = {}
if n > Inputs and n <= MaxNodes then
cell.x = 140
cell.y = 40
cell.value = neuron.value
-- if neuron.value >0 then
-- neuron.switcher = true
-- end
cells[n] = cell
end
end
local genes = {}
if nswitch then
genes = genome.genes
else
genes = genome.genes2
end
for n=1,4 do
for _,gene in pairs(genes) do
if gene.enabled then
local c1 = cells[gene.into]
local c2 = cells[gene.out]
if gene.into > Inputs and gene.into <= MaxNodes then
c1.x = 0.75*c1.x + 0.25*c2.x
if c1.x >= c2.x then
c1.x = c1.x - 40
end
if c1.x < 90 then
c1.x = 90
end
if c1.x > 220 then
c1.x = 220
end
c1.y = 0.75*c1.y + 0.25*c2.y
end
if gene.out > Inputs and gene.out <= MaxNodes then
c2.x = 0.25*c1.x + 0.75*c2.x
if c1.x >= c2.x then
c2.x = c2.x + 40
end
if c2.x < 90 then
c2.x = 90
end
if c2.x > 220 then
c2.x = 220
end
c2.y = 0.25*c1.y + 0.75*c2.y
end
end
end
end
gui.drawbox(50-BoxRadius*5-3,70-BoxRadius*5-3,50+BoxRadius*5+2,70+BoxRadius*5+2, toRGBA(0x80808080), toRGBA(0xFF000000))
for n,cell in pairs(cells) do
if n > Inputs or cell.value ~= 0 then
local color = math.floor((cell.value+1)/2*256)
if color > 255 then color = 255 end
if color < 0 then color = 0 end
local opacity = 0xFF000000
if cell.value == 0 then
opacity = 0x50000000
end
inversecolor = (color - 255) * (-1)
bordercolor = 0x50000000 + inversecolor*0x10000 + inversecolor*0x100 + inversecolor
color = opacity + color*0x10000 + color*0x100 + color
gui.drawbox(cell.x-2,cell.y-2,cell.x+2,cell.y+2,toRGBA(color), toRGBA(bordercolor))
end
end
for _,gene in pairs(genes) do
if gene.enabled then
local c1 = cells[gene.into]
local c2 = cells[gene.out]
local opacity = 0xF0000000
if pool.generation > 50 then
opacity = 0x80000000
end
if c1.value == 0 then
if pool.generation <10 then --slowly reducing the visibility of non active connections
opacity = 0x80000000
elseif pool.generation < 15 then
opacity = 0x60000000
elseif pool.generation < 25 then
opacity = 0x40000000
end
end
local color = 0x80-math.floor(math.abs(sigmoid(gene.weight))*0x80)
if gene.weight > 0 then
color = opacity + 0x8000 + 0x10000*color
else
color = opacity + 0x800000 + 0x100*color
end
if c1.value ~= 0 or pool.generation < 5 then -- remove non active connections from generation 50 onwards
gui.drawline(c1.x+1, c1.y, c2.x-3, c2.y, toRGBA(color))
end
end
end
gui.drawbox(49,71,51,78,toRGBA(0x80FF0000),toRGBA(0x00000000))
if mutation_rates_disp then
local pos = 120
for mutation,rate in pairs(genome.mutationRates) do
gui.drawtext(16, pos, mutation .. ": " .. rate, toRGBA(0xFF000000), 0x0)
pos = pos + 8
end
end
end
function writeFile(filename)
local file = io.open(filename, "w")
file:write(pool.generation .. "\n")
file:write(pool.maxFitness .. "\n")
file:write(Population .. "\n")
file:write(pool.innovation .. "\n")
file:write(bottleneckcounter .. "\n")
file:write(#pool.species .. "\n")
for n,species in pairs(pool.species) do
file:write(species.topFitness .. "\n")
file:write(species.staleness .. "\n")
file:write(#species.genomes .. "\n")
for m,genome in pairs(species.genomes) do
file:write(genome.fitness .. "\n")
file:write(genome.maxneuron .. "\n")
for mutation,rate in pairs(genome.mutationRates) do
file:write(mutation .. "\n")
file:write(rate .. "\n")
end
file:write("done\n")
file:write(#genome.genes .. "\n")
for l,gene in pairs(genome.genes) do
file:write(gene.into .. " ")
file:write(gene.out .. " ")
file:write(gene.weight .. " ")
file:write(gene.innovation .. " ")
if(gene.enabled) then
file:write("1\n")
else
file:write("0\n")
end
end
file:write(#genome.genes2 .. "\n")
for l,gene in pairs(genome.genes2) do
file:write(gene.into .. " ")
file:write(gene.out .. " ")
file:write(gene.weight .. " ")
file:write(gene.innovation .. " ")
if(gene.enabled) then
file:write("1\n")
else
file:write("0\n")
end
end
file:write(genome.networkswitch .. "\n")
file:write(#genome.oscilations .. "\n")
for i=1,#genome.oscilations do
file:write(genome.oscilations[i] .. "\n")
end
end
end
file:close()
end
function savePool() --used to save when a new max fitness is reached
local filename = "backups/backup." .. pool.generation .. "." .. SAVE_LOAD_FILE
writeFile(filename)
emu.print("saved pool due to new max fitness") --suggestion by DeltaLeeds
end
function loadFile(filename)
local file = io.open(filename, "r")
pool = newPool()
pool.generation = file:read("*number")
pool.maxFitness = file:read("*number")
Population = file:read("*number")
pool.innovation = file:read("*number")
bottleneckcounter = file:read("*number")
local numSpecies = file:read("*number")
for s=1,numSpecies do
local species = newSpecies()
table.insert(pool.species, species)
species.topFitness = file:read("*number")
species.staleness = file:read("*number")
local numGenomes = file:read("*number")
for g=1,numGenomes do
local genome = newGenome()
table.insert(species.genomes, genome)
genome.fitness = file:read("*number")
genome.maxneuron = file:read("*number")
local line = file:read("*line")
while line ~= "done" do
genome.mutationRates[line] = file:read("*number")
line = file:read("*line")
end
local numGenes = file:read("*number")
for n=1,numGenes do
local gene = newGene()
table.insert(genome.genes, gene)
local enabled
--I am able to revert back to the original loading script with FCEUX
gene.into, gene.out, gene.weight, gene.innovation, enabled = file:read("*number", "*number", "*number", "*number", "*number")
if enabled == 0 then
gene.enabled = false
else
gene.enabled = true
end
end
local numGenes2 = file:read("*number")
for n=1,numGenes2 do
local gene2 = newGene()
table.insert(genome.genes2, gene2)
local enabled
--I am able to revert back to the original loading script with FCEUX
gene2.into, gene2.out, gene2.weight, gene2.innovation, enabled = file:read("*number", "*number", "*number", "*number", "*number")
if enabled == 0 then
gene2.enabled = false
else
gene2.enabled = true
end
end
genome.networkswitch = file:read("*number")
local numosci = file:read("*number")
for i=1,numosci do
genome.oscilations[i] = file:read("*number")
end
end
end
file:close()
while fitnessAlreadyMeasured() do
nextGenome()
end
initializeRun()
pool.currentFrame = pool.currentFrame + 1
end
function playTop()
local maxfitness = 0
local maxs, maxg
for s,species in pairs(pool.species) do
for g,genome in pairs(species.genomes) do
if genome.fitness > maxfitness then
maxfitness = genome.fitness
maxs = s
maxg = g
end
end
end
pool.currentSpecies = maxs
pool.currentGenome = maxg
pool.maxFitness = maxfitness
initializeRun()
pool.currentFrame = pool.currentFrame + 1
return
end
function fitnesstracker()
if pool.maxFitness > 0 then
local file = io.open('fitnesstracker.txt', "a")
file:write("Gen: " .. pool.generation .. " Species: " .. pool.currentSpecies .. " Genome: " .. pool.currentGenome .. " Fitness: ".. pool.maxFitness .. "\n")
file:close()
end
end
function writelatestgen() --used to write the generation tracker to a file
local file = io.open('backups/latestgen', "w")
file:write(SAVE_LOAD_FILE .. "\n")
file:write(savestate_slot .. "\n")
file:write(pool.generation)
file:close()
end
function readlatestgen() --used to retrieve the latest generation from a file
local file = io.open('backups/latestgen', "r")
SAVE_LOAD_FILE = file:read("*line")
savestate_slot = file:read("*number")
SavestateObj = savestate.object(savestate_slot)
local latestgen = file:read("*number")
file:close()
local name = 'backups/backup.'.. latestgen .. '.' .. SAVE_LOAD_FILE
print('loaded: '.. name)
return name
end
function keyboardinput()
local keyboard = input.get()
if keyboard['N'] and keyflag == false then
if neural_net_disp then
neural_net_disp = false
else
neural_net_disp = true
end
keyflag = true
end
if keyboard['M'] and keyflag == false then
if mutation_rates_disp then
mutation_rates_disp = false
else
mutation_rates_disp = true
end
keyflag = true
end
if keyboard['L'] and keyflag == false then
if not nopopup then
local loadyn = input.popup('Are you sure you want to load the last saved generation?')
end
if loadyn == 'yes' or nopopup then
name = readlatestgen()
loadFile(name)
end
keyflag = true
end
if keyboard['B'] and keyflag == false then
if not nopopup then
local loadyn = input.popup('Are you sure you want to show the topFitness genome?')
end
if loadyn == 'yes' or nopopup then
playTop()
end
end
if keyboard['O'] and keyflag == false then
pcall(InsertGenome)
keyflag = true
end
if keyboard['U'] and keyflag == false then
pcall(PlaceGenome)
keyflag = true
end
if keyboard['S'] and keyflag == false then
savePool()
keyflag = true
end
if not (keyboard['N'] or keyboard['L'] or keyboard['M'] or keyboard['T'] or keyboard['B'] or keyboard['S']) and keyflag == true then
keyflag = false
end
if keyboard['T'] and keyflag == false then
if turbo then
emu.speedmode("normal")
turbo = false
print('stop turbo')
elseif not turbo and not restoreturbo then
emu.speedmode("turbo")
turbo = true
elseif restoreturbo then
restoreturbo = false
end
keyflag = true
end
end
if savedpool then
loadFile(savedpool)
end
writeFile("temp.pool")
requiredFitness = 0
if SAVE_LOAD_FILE == nil then
SAVE_LOAD_FILE = "SMB".. CurrentWorld+1 .."-".. CurrentLevel ..".state.pool"
end
while true do
keyboardinput()
local backgroundColor = toRGBA(0xD0FFFFFF)
gui.drawbox(0, 0, 260, 30, backgroundColor, backgroundColor)
if pool.maxcounter == nil then
pool.maxcounter = 0
emu.print("caught missing variable")
end
local species = pool.species[pool.currentSpecies]
local genome = species.genomes[pool.currentGenome]
if neural_net_disp then
displayGenome(genome)
end
if pool.generation == 0 and pool.currentSpecies ~= 1 and not turbo and AllowAutoTurbo and FirstGenSpeedup then
emu.speedmode("turbo")
turbo = true
end
if genome.networkswitch>0 then
if math.floor(rightmost/10 / genome.networkswitch) % 50 == 0 and ntrigger == false then
if nswitch == true then
nswitch = false
else
nswitch = true
end
ntrigger = true
elseif math.floor(rightmost/10 / genome.networkswitch) % 50 ~= 0 and ntrigger == true then
ntrigger = false
end
end
if pool.currentFrame%5 == 0 then
evaluateCurrent()
end
joypad.set(player, controller)
getPositions()
if mariostate == 2 or mariostate == 3 then --detects when mario enters a pipe
marioPipeEnter = true
if mariostate == 2 then
adjustfitness = true
else
adjustfitness = false
end
end
if (CurrentWorld == 3 and CurrentLevel == 4) or (CurrentWorld == 6 and CurrentLevel == 4) or (CurrentWorld == 7 and CurrentLevel == 3) then
if currentTracker == currentscreen or currentTracker == nextscreen and not mariopipe then --follows the progress of mario
currentTracker = nextscreen
elseif currentTracker > nextscreen and not mariopipe and not marioPipeEnter and mariostate ~= 7 then --if mario suddenly goes back trigger loop detection
loop = true
elseif currentscreen == 0 and nextscreen == 0 and mariopipe and mariostate ~= 7 then -- if mario enters a piperoom negate loop detection
loop = false
currentTracker = nextscreen
elseif nextscreen >= (currentTracker - 1) and mariopipe and mariostate ~= 7 then -- if mario goes forward in the level by entering a pipe negate loop detection
loop = false
currentTracker = nextscreen
pipeexit = true
else --if nothing is true then we are in a loop
if not pipeexit then --prevents the extra one frame flicker of the tracker
loop = true
else
pipeexit = false
end
end
end
if marioPipeEnter == true and mariostate == 7 then --set mariopipe when exiting the pipe
mariopipe = true
end
if adjustfitness then
if mariopipe == true and mariostate ~= 7 and rightmost > marioX and not mariohole then --calculate offset
offset = rightmost - marioX
mariopipe = false
marioPipeEnter = false
elseif mariopipe == true and mariostate ~= 7 and rightmost <= marioX and not mariohole then --if the exit coordinates are higher than rightmost after exiting a pipe set offset to 0
offset = 0
mariopipe = false
marioPipeEnter = false
end
else
if mariopipe == true and mariostate ~= 7 and not mariohole then --calculate offset --and rightmost > marioX
offset = rightmost - marioX
mariopipe = false
marioPipeEnter = false
end
end
if marioY + memory.readbyte(0xB5)*255 > 512 then --added the fallen into a hole variable that will activate when mario goes below screen
mariohole = true
end
if marioX > 2370 and marioX < 3000 and (CurrentWorld == 7 and CurrentLevel == 3) then
if marioX < 2460 then
if 2370 + offset - marioY + 192 + (50 *memory.readbyte(0x06D9)) > rightmost and not mariohole and not loop and not marioPipeEnter and not killtrigger and not falling then
rightmost = 2370 + offset - marioY + 192 + (50 *memory.readbyte(0x06D9))
timebonus = timebonus + (TimeoutConstant - timeout) * 11/20
timeout = TimeoutConstant
end
end
else
if marioX + offset - marioY/5 + 192/5 + (50 *memory.readbyte(0x06D9)) > rightmost and not mariohole and not loop and not marioPipeEnter and not killtrigger and not falling then
rightmost = marioX + offset - marioY/5 + 192/5 + (50 *memory.readbyte(0x06D9))
timebonus = timebonus + (TimeoutConstant - timeout) * 11/20
timeout = TimeoutConstant
end
end
if marioPipeEnter == true then --freeze fitness and timer when mario enters a pipe
timeout = timeout + 1
timebonus = timebonus + 2/3
end
if (mariostate == 4 or (memory.readbyte(0x000C) == 1 and statuscheck == 0) or ((memory.readbyte(0x06D6) ~= 0 and memory.readbyte(0x06D6) ~= 1) and marioPipeEnter)) or endlevel then --and statuscheck == 0
timeout = timeout +1
timebonus = timebonus + 2/3
endlevel = true
end
if endlevel and mariostate == 8 and demoruncheck == 1 and statuscheck == 3 then
local timeoutBonus = pool.currentFrame / 4
local fitness = rightmost - pool.currentFrame / 2 +40 + timebonus
genome.fitness = fitness
pool.maxFitness = fitness
fitnesstracker()
savePool()
if not rerunning then
SAVE_LOAD_FILE = "SMB".. CurrentWorld+1 .."-".. CurrentLevel ..".state.pool"
previous_savestate = savestate_slot
savestate_slot = savestate_slot + 1
if savestate_slot == 10 then
savestate_slot = 1
end
Savestatebackup = savestate.object(savestate_slot)
savestate.save(Savestatebackup)
savestate.persist(Savestatebackup)
savestate.save(SavestateObj)
killcounter = 0
end
SavestateObj = savestate.object(savestate_slot)
endlevel = false
if (turbo or restoreturbo) and not rerunning then
rerunning = true
requiredFitness = 0
emu.speedmode("normal")
turbo = false
restoreturbo = false
reRun()
else
rerunning = false
preparenext = true
end
end
if CurrentWorld == 0 and CurrentLevel == 0 and demoruncheck == 0 then
savestate.load(SavestateObj)
end
timeout = timeout - 1
if preparenext then
rerunning = false
preparenext = false
GenerateNewKeepBest(genome)
end
if rightmost > maxright then
maxright = rightmost
end
if rightmost > maxright*0.85 and AllowSlowdownNearMax then
if turbo and maxright > 200 and AllowSlowdownNearMax and not restoreturbo then
emu.speedmode("normal")
turbo = false
restoreturbo = true
end
elseif species.turbo == 'off' then
if turbo then
emu.speedmode("normal")
turbo = false
end
elseif requiredFitness > rightmost and requiredFitness > 10 and not turbo then
emu.speedmode("turbo")
turbo = true
restoreturbo = false
elseif requiredFitness < rightmost and turbo and requiredFitness > 10 then
emu.speedmode("normal")
turbo = false
restoreturbo = true
elseif restoreturbo and not (requiredFitness > 10) then
emu.speedmode("turbo")
turbo = true
restoreturbo = false
end
local timeoutBonus = pool.currentFrame / 4
if (mariostate == 11 or mariohole or math.floor(rightmost - (pool.currentFrame) / 2 + - (timeout + timeoutBonus)*2/3 +40 + timebonus) < -50 ) and timeout > 40-timeoutBonus then
local temptimeout = timeout
timeout = 30 - timeoutBonus
local difference = temptimeout - timeout
pool.currentFrame = pool.currentFrame + difference
killtrigger = true
end
local timeoutBonus = pool.currentFrame / 4
if timeout + timeoutBonus <= 0 then
local fitness = rightmost - pool.currentFrame / 2 +40 + timebonus
if fitness == 0 then
fitness = -1
end
genome.fitness = fitness
if fitness > pool.maxFitness then
pool.maxFitness = fitness
fitnesstracker()
savePool()
end
if fitness > pool.maxFitness - 30 then
pool.maxcounter = pool.maxcounter + 1
end
emu.print("Gen " .. pool.generation .. " species " .. pool.currentSpecies .. " genome " .. pool.currentGenome .. " fitness: " .. fitness)
pool.currentSpecies = 1
pool.currentGenome = 1
while fitnessAlreadyMeasured() do
offset = 0 --offset in x coordinates for pipe
timebonus = 0 --used to freeze fitness
mariohole = false --reset the fallen into hole variable
mariopipe = false --is mario exiting a pipe or just teleporting
loop = false --is this a loop?
marioPipeEnter = false --has mario entered a pipe
currentTracker = 1 --tracker for the loop
killtrigger = false
ntrigger = true
nswitch = true
if rerunning then
rerunning = false
preparenext = true
end
nextGenome()
end
initializeRun()
end
local measured = 0
local total = 0
for _,species in pairs(pool.species) do
for _,genome in pairs(species.genomes) do
total = total + 1
if genome.fitness ~= 0 then
measured = measured + 1
end
end
end
gui.opacity(1) --made the top banner opaque. you can change this if you want a slightly transparent banner
if turbo then
gui.drawtext(250, 11, "T", toRGBA(0xFFFF0000), 0x0)
end
gui.drawtext(5, 11, "Gen " .. pool.generation .. " species " .. pool.currentSpecies .. " genome " .. pool.currentGenome .. " (" .. math.floor(measured/total*100) .. "%) ".. "Pop: ".. Population, toRGBA(0xFF000000), 0x0)
gui.drawtext(5, 20, "Fitness: " .. math.floor(rightmost - (pool.currentFrame) / 2 + - (timeout + timeoutBonus)*2/3 +40 + timebonus) , toRGBA(0xFF000000), 0x0) --the (timeout + timeoutBonus)*2/3 makes sure that the displayed fitness remains stable for the viewers pleasure
gui.drawtext(80, 20, "Max Fitness:" .. math.floor(pool.maxFitness) .. " (".. math.floor(secondbest) .. ")".. " ".. "(" ..pool.maxcounter .. "x)", toRGBA(0xFF000000), 0x0)
pool.currentFrame = pool.currentFrame + 1
emu.frameadvance();
end
Comments
0 B
|👍
/👎