async Task ReceiveClientTry(TryRequest t) // Simulate the POST /try function
{
WriteLine($"..... receive try {t}");
await Task.Delay(0);
GeneticAlgorithm(t);
return; // emphatic empty return
}
async void GeneticAlgorithm(TryRequest treq)
{
WriteLine($"..... GeneticAlgorithm {treq}");
await Task.Delay(0);
// just an ad-hoc PR test - you will remove this
// await ProportionalRandomTest ();
// YOU CODE GOES HERE
// FOLLOW THE GIVEN PSEUDOCODE
// ...
var id = treq.id;
var monkeys = treq.monkeys;
if (monkeys % 2 != 0)
{
monkeys += 1;
}
var length = treq.length;
var crossover = treq.crossover / 100.0;
var mutation = treq.mutation / 100.0;
var limit = treq.limit;
if (limit == 0)
{
limit = 1000;
}
// ... UDOO
for (int loop = 0; loop < limit; loop++)
{
// ... UDOO
}
}
async void GeneticAlgorithm(TryRequest treq)
{
WriteLine($"..... GeneticAlgorithm {treq}");
await Task.Delay(0);
var id = treq.id;
var monkeys = treq.monkeys;
if (monkeys % 2 != 0)
{
monkeys += 1;
}
var length = 0;
if (treq.length == 0)
{
var nullText = new[] { "" }.ToList();
var r = new AssessRequest {
id = id, genomes = nullText
};
AssessResponse ares = await PostFitnessAssess(r);
var scores = ares.scores;
length = scores[0];
}
else
{
length = treq.length;
}
var crossover = treq.crossover / 100.0;
var mutation = treq.mutation / 100.0;
var limit = treq.limit;
if (limit == 0)
{
limit = 1000;
}
var topscore = int.MaxValue;
var genomes = Enumerable.Range(0, monkeys).Select(i => {
var genome = Enumerable.Range(0, length).Select(n => {
return((char)NextInt(32, 127));
}).ToList();
return(string.Join("", genome));
}).ToList();
var obj1 = new AssessRequest {
id = id, genomes = genomes
};
for (int loop = 0; loop < limit; loop++)
{
AssessResponse x = await PostFitnessAssess(obj1);
var Id = x.id;
var scores = x.scores;
var smallest = scores.Min();
var largest = scores.Max();
var index = scores.IndexOf(smallest);
var genome_list = obj1.genomes;
var genome = genome_list[index];
if (smallest < topscore)
{
var obj2 = new TopRequest {
id = Id, loop = loop, score = smallest, genome = genome
};
await PostClientTop(obj2);
topscore = smallest;
}
if (smallest == 0)
{
break;
}
else
{
var weights = scores.Select(s => {
return(largest - s + 1);
});
var para = treq.parallel;
if (para)
{
var newGenomes = ParallelEnumerable.Range(1, monkeys / 2).SelectMany <int, string>(i => {
var c1 = "";
var c2 = "";
var index1 = ProportionalRandom(weights.ToArray(), weights.Sum());
var index2 = ProportionalRandom(weights.ToArray(), weights.Sum());
var p1 = genome_list[index1];
var p2 = genome_list[index2];
if (NextDouble() < crossover)
{
var crossoverIndex = NextInt(0, length);
c1 = p1.Substring(0, crossoverIndex) + p2.Substring(crossoverIndex, p2.Length - crossoverIndex);
c2 = p2.Substring(0, crossoverIndex) + p1.Substring(crossoverIndex, p1.Length - crossoverIndex);
}
else
{
c1 = p1;
c2 = p2;
}
if (NextDouble() < mutation)
{
char[] c = c1.ToCharArray();
c[NextInt(0, c1.Length)] = (char)NextInt(32, 127);
c1 = new string(c);
}
if (NextDouble() < mutation)
{
char[] c = c2.ToCharArray();
c[NextInt(0, c2.Length)] = (char)NextInt(32, 127);
c2 = new string(c);
}
var mylist = new List <string>();
mylist.Add(c1);
mylist.Add(c2);
return(mylist);
}).ToList();
obj1.id = Id;
obj1.genomes = newGenomes;
}
else
{
var newGenomes = Enumerable.Range(1, monkeys / 2).SelectMany <int, string>(i => {
var c1 = "";
var c2 = "";
var index1 = ProportionalRandom(weights.ToArray(), weights.Sum());
var index2 = ProportionalRandom(weights.ToArray(), weights.Sum());
var p1 = genome_list[index1];
var p2 = genome_list[index2];
if (NextDouble() < crossover)
{
var crossoverIndex = NextInt(0, length);
c1 = p1.Substring(0, crossoverIndex) + p2.Substring(crossoverIndex, p2.Length - crossoverIndex);
c2 = p2.Substring(0, crossoverIndex) + p1.Substring(crossoverIndex, p1.Length - crossoverIndex);
}
else
{
c1 = p1;
c2 = p2;
}
if (NextDouble() < mutation)
{
char[] c = c1.ToCharArray();
c[NextInt(0, c1.Length)] = (char)NextInt(32, 127);
c1 = new string(c);
}
if (NextDouble() < mutation)
{
char[] c = c2.ToCharArray();
c[NextInt(0, c2.Length)] = (char)NextInt(32, 127);
c2 = new string(c);
}
var mylist = new List <string>();
mylist.Add(c1);
mylist.Add(c2);
return(mylist);
}).ToList();
obj1.id = Id;
obj1.genomes = newGenomes;
}
}
}
}
async void GeneticAlgorithm(TryRequest treq)
{
var length = treq.length;
var monkeys = treq.monkeys;
List <string> post = new List <string>();
string bestStr = "";
if (monkeys % 2 != 0)
{
monkeys += 1;
}
//discover length
if (length == 0)
{
List <string> test = new List <string>()
{
""
};
var content = new StringContent(JsonSerializer.Serialize(test), System.Text.Encoding.UTF8,
"application/json");
var response = await client.PostAsync("http://localhost:8091/assess", content);
var l = await response.Content.ReadAsAsync <List <int> >();
length = l[0];
}
//create start genome
IEnumerable <string> strings =
Enumerable.Repeat(
Nextstr(length), monkeys);
post = strings.ToList();
WriteLine($"..... POST length {length}");
for (var count = 0; count < treq.limit; count++)
{
//Send content
var content = new StringContent(JsonSerializer.Serialize(post), System.Text.Encoding.UTF8,
"application/json");
var response = await client.PostAsync("http://localhost:8091/assess", content);
var ress = await response.Content.ReadAsAsync <List <int> >();
//find best
var min = ress.Min();
var tempBestString = post[ress.IndexOf(min)];
var tempBest = min;
if (0 == tempBest)
{
best = tempBest;
bestStr = tempBestString;
var top = new TopRequest(8081, count, best, bestStr);
var topcont = new StringContent(JsonSerializer.Serialize(top), System.Text.Encoding.UTF8,
"application/json");
var resp = await client.PostAsync($"http://localhost:{treq.id}/top", topcont);
WriteLine($" DONE {tempBestString}");
return;
}
if (tempBest < best)
{
best = tempBest;
bestStr = tempBestString;
WriteLine($" BEST {tempBestString}");
var top = new TopRequest(8081, count, best, bestStr);
var topcont = new StringContent(JsonSerializer.Serialize(top), System.Text.Encoding.UTF8,
"application/json");
var resp = await client.PostAsync("http://localhost:8101/top", topcont);
}
//if its parallel
if (treq.parallel)
{
post = ParallelEnumerable.Range(1, monkeys / 2)
.SelectMany <int, string>(i =>
{
var p1 = post[ProportionalRandom(ress.ToArray(), ress.Sum())];
var p2 = post[ProportionalRandom(ress.ToArray(), ress.Sum())];
var c1 = "";
var c2 = "";
//cross over chance
if (NextInt(0, 100) < treq.crossover)
{
var Index = NextInt(0, p1.Length - 1);
c1 = p1.Substring(0, Index) + p2.Substring(Index, p2.Length - Index);
c2 = p2.Substring(0, Index) + p1.Substring(Index, p1.Length - Index);
}
else
{
c1 = p1;
c2 = p2;
}
if (NextInt(0, 100) < treq.mutation)
{
var item = NextInt(0, c1.Length);
StringBuilder strBuilder = new System.Text.StringBuilder(c1);
strBuilder[item] = Convert.ToChar(NextInt(32, 126));
;
c1 = strBuilder.ToString();
}
if (NextInt(0, 100) < treq.mutation)
{
var item = NextInt(0, c2.Length);
StringBuilder strBuilder = new System.Text.StringBuilder(c2);
strBuilder[item] = Convert.ToChar(NextInt(32, 126));
;
c2 = strBuilder.ToString();
}
return(new[] { c1, c2 });
}).ToList();
}
else
{
post = Enumerable.Range(1, monkeys / 2)
.SelectMany <int, string>(i =>
{
var p1 = post[ProportionalRandom(ress.ToArray(), ress.Sum())];
var p2 = post[ProportionalRandom(ress.ToArray(), ress.Sum())];
var c1 = "";
var c2 = "";
//cross over chance
if (NextInt(0, 100) < treq.crossover)
{
var Index = NextInt(0, p1.Length - 1);
c1 = p1.Substring(0, Index) + p2.Substring(Index, p2.Length - Index);
c2 = p2.Substring(0, Index) + p1.Substring(Index, p1.Length - Index);
}
else
{
c1 = p1;
c2 = p2;
}
if (NextInt(0, 100) < treq.mutation)
{
var item = NextInt(0, c1.Length);
StringBuilder strBuilder = new System.Text.StringBuilder(c1);
strBuilder[item] = Convert.ToChar(NextInt(32, 126));
;
c1 = strBuilder.ToString();
}
if (NextInt(0, 100) < treq.mutation)
{
var item = NextInt(0, c2.Length);
StringBuilder strBuilder = new System.Text.StringBuilder(c2);
strBuilder[item] = Convert.ToChar(NextInt(32, 126));
;
c2 = strBuilder.ToString();
}
return(new[] { c1, c2 });
}).ToList();
}
}
}
async void GeneticAlgorithm(TryRequest treq)
{
WriteLine($"..... GeneticAlgorithm {treq}");
await Task.Delay(0);
var id = treq.id;
var monkeys = treq.monkeys;
if (monkeys % 2 != 0)
{
monkeys += 1;
}
var length = treq.length;
var crossover = treq.crossover / 100.0;
var mutation = treq.mutation / 100.0;
var limit = treq.limit;
if (limit == 0)
{
limit = 1000;
}
var isParallel = treq.parallel;
if (treq.length == 0) // Dynamic cal length
{
var list = new List <string>()
{
""
};
var res = await PostFitnessAssess(new AssessRequest { id = id, genomes = list, });
length = res.scores[0];
}
// Create randoms
currentGeneration = new List <string>();
currentGeneration = createRandoms(monkeys, length);
string currentBestGenome = currentGeneration[0]; // Randomly choose initially
string targettxt = currentBestGenome;
for (int loop = 0; loop < limit; loop++) // The evolutaion loop
// Get the fitness values from the fitness server
{
var res = await PostFitnessAssess(new AssessRequest { id = id, genomes = currentGeneration, });
var fs = res.scores;
currentBestGenome = getCurrentBest(fs);
// Get the fitness of current best as well as the targettext
var list = new List <string>()
{
currentBestGenome, targettxt
};
var res2 = await PostFitnessAssess(new AssessRequest { id = id, genomes = list, });
var fs2 = res2.scores;
if (fs2[0] < fs2[1]) // current best genome's fitness < best's fitness
{
targettxt = currentBestGenome;
// Post the evolutaion string to the clinet
await PostClientTop(new TopRequest { id = id, loop = loop, score = fs2[0], genome = targettxt, });
if (fs2[0] == 0)
{
break; // Find the target and break the loop
}
}
// Create new generations
currentGeneration = createNewGeneration(monkeys, mutation, crossover, fs, isParallel);
}
}
