async Task <AssessResponse> PostFitnessAssess(AssessRequest areq)
{
// ... UDOO
await Task.Delay(0);
var ares = new AssessResponse();
return(ares);
}
async Task <AssessResponse> PostFitnessAssess(AssessRequest a)
{
await Task.Delay(0);
var client = new HttpClient();
client.BaseAddress = new Uri("http://localhost:8091/");
client.DefaultRequestHeaders.Accept.Clear();
client.DefaultRequestHeaders.Accept.Add(
new MediaTypeWithQualityHeaderValue("application/json"));
var hrm = await client.PostAsJsonAsync("/assess", a);
hrm.EnsureSuccessStatusCode();
var ares = await hrm.Content.ReadAsAsync <AssessResponse>();
return(ares);
}
string target = ""; // simplified Fitness stuff, shouldn't be here
async Task <AssessResponse> PostFitnessAssess(AssessRequest areq)
{
// fake Fitness post request&response
// replace this code by actual POST /assess request&response
// var client = new HttpClient ();
// ...
await Task.Delay(0);
var scores = areq.genomes.Select(g => {
var len = Math.Min(target.Length, g.Length);
var h = Enumerable.Range(0, len)
.Sum(i => Convert.ToInt32(target[i] != g[i]));
h = h + Math.Max(target.Length, g.Length) - len;
return(h);
}).ToList();
return(new AssessResponse {
id = areq.id, scores = scores
});
}
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;
}
}
}
}