public double Error(double[][] x, double[][] y, IList <IRule> ruleBase)
{
if (x.Length != y.Length)
{
throw new Exception("Input and desired output lengths not match");
}
if (ruleBase == null || ruleBase.Count == 0)
{
throw new Exception("Incorrect rulebase");
}
int outputDim = ruleBase[0].Z.Length;
int numOfRules = ruleBase.Count;
double globalError = 0.0;
for (int sample = 0; sample < x.Length; sample++)
{
double[] o = ANFIS.Inference(x[sample], ruleBase);
for (int C = 0; C < outputDim; C++)
{
globalError += Math.Abs(o[C] - y[sample][C]);
}
}
return(globalError / x.Length);
}
private static void subtestLogisticsMap <T>(double[][] x, double[][] y, double[][] tx, double[][] ty, ITraining bprop) where T : IRule, new()
{
KMEANSExtractorIO extractor = new KMEANSExtractorIO(10);
var timer = Stopwatch.StartNew();
ANFIS fis = ANFISBuilder <T> .Build(x, y, extractor, bprop, 1000);
timer.Stop();
double err = bprop.Error(tx, ty, fis.RuleBase);
Trace.WriteLine(string.Format("[{1} - {4}]\tLogistic map Error {0}\tElapsed {2}\tRuleBase {3}", err, bprop.GetType().Name, timer.Elapsed, fis.RuleBase.Length, typeof(T).Name), "training");
Assert.IsFalse(err > 1e-2);
}
public void ANFIS_OutputFromDataSet()
{
var sampleSize = RobotArmDataSet.Input.Length - 1;
//StochasticBatch sprop = new StochasticBatch(sampleSize, 1e-5);
//sprop.UnknownCaseFaced += AddRule<GaussianRule2>;
//var sprop = new Backprop(1e-1);
var sprop = new StochasticQprop(sampleSize);
var extractor = new KMEANSExtractorIO(25);
//ANFIS fis = ANFISBuilder<GaussianRule2>.Build(RobotArmDataSet.Input, RobotArmDataSet.OutputTheta1, extractor, sprop, 150);
ANFIS fis = ANFISBuilder <GaussianRule> .Build(RobotArmDataSet.Input, RobotArmDataSet.OutputTheta1, extractor, sprop, 150);
var output1 = fis.Inference(new[] { 1.10413546487088, 2.81104319371924 }).FirstOrDefault(); // 1.1
var output2 = fis.Inference(new[] { 2.31665592712393, 1.9375717475909 }).FirstOrDefault(); // 0.6
var output3 = fis.Inference(new[] { 2.88944142930409, 16.7526454098038 }).FirstOrDefault(); // 1.4
}
private static void Main(string[] args)
{
const int numberOfRules = 10;
var anfis = new ANFIS.ANFIS(numberOfRules, 10_000);
var gradSus = anfis.Run(useBatch: false, eta: 0.001, etaZ: 0.01, verbose: false);
// var anfisBatch = new ANFIS(numberOfRules, 10_000);
// var gradSus = anfisBatch.Run(useBatch: true, batchSize: 4, eta: 0.0001, etaZ: 0.01, verbose: false);
// var samples = ANFIS.GenerateSamples();
// const string root = "C:/Faks/NENR/NENR6/NENR6/Data/";
// var testRoot = root + $"{numberOfRules}.txt";
// var testRoot = root + "sampled.txt";
// FileHelper.WriteToFile(samples, testRoot);
// WriteDifferencesToFile();
}
public static void WriteDifferencesToFile()
{
const int numberOfRules = 10;
var anfis = new ANFIS.ANFIS(numberOfRules, 10_000);
var gradSus = anfis.Run(useBatch: false, eta: 0.001, etaZ: 0.01, verbose: false);
// var anfisBatch = new ANFIS(numberOfRules, 10_000);
// var gradSus = anfisBatch.Run(useBatch: true, batchSize: 4, eta: 0.0001, etaZ: 0.01, verbose: false);
var samples = ANFIS.ANFIS.GenerateSamples();
const string root = "C:/Faks/NENR/NENR6/NENR6/Data/SGD/Differences/";
var testRoot = root + $"{numberOfRules}.txt";
var sampleDifferences = samples;
for (var i = 0; i < samples.Count; i++)
{
sampleDifferences[i] = new Sample(samples[i].X, samples[i].Y, samples[i].Z - gradSus[i].Z);
}
FileHelper.WriteToFile(sampleDifferences, testRoot);
}
void Awake()
{
//Check if there is already an instance of SoundManager
if (instance == null)
{
//if not, set it to this.
instance = this;
}
//If instance already exists:
else if (instance != this)
{
//Destroy this, this enforces our singleton pattern so there can only be one instance of SoundManager.
Destroy(gameObject);
}
//Set SoundManager to DontDestroyOnLoad so that it won't be destroyed when reloading our scene.
DontDestroyOnLoad(gameObject);
Backprop bprop = new Backprop(1e-2);
KMEANSExtractorIO extractor = new KMEANSExtractorIO(5);
fis = ANFISBuilder <GaussianRule> .Build(gameParams, musicParams, extractor, bprop, 5);
}
private static void ReportTwo(List <UserData> userRawData, Dictionary <int, string> juxtaposition)
{
var ruleset = new List <IRule>
{
new SimpleLinearRule(
new double[] { -0.1, -1 },
new double[] { 0.1, 150 },
0),
new SimpleLinearRule(
new double[] { 0, -1 },
new double[] { 3, 2 },
1),
new SimpleLinearRule(
new double[] { 1, -1 },
new double[] { 6, 3 },
2),
new SimpleLinearRule(
new double[] { 1, -1 },
new double[] { 11, 4 },
3),
new SimpleLinearRule(
new double[] { 0, 0 },
new double[] { 150, 150 },
4),
new SimpleLinearRule(
new double[] { 2, -1 },
new double[] { 150, 4 },
5),
new SimpleLinearRule(
new double[] { 2, -1 },
new double[] { 21, 3 },
6),
};
var fis = new ANFIS(ruleset);
var groopy = userRawData.ToLookup(z => z.Id, z => z.GroupId).Where(z => z.Count() < 150);
var sb = new StringBuilder();
sb.AppendLine($"ID, {string.Join(", ", Enumerable.Range(0, 7).Select(z => $"Стадия {z}"))}");
var count = groopy.Count();
foreach (var user in groopy)
{
var codes = user.Where(z => juxtaposition.ContainsKey(z)).Select(z => juxtaposition[z]);
if (codes.Any())
{
var line = $"{user.Key}, ";
var input = new double[]
{
codes.Count(z => z.Equals("s6-2")),
codes.Count(z => z.Equals("s6-3"))
};
var res = fis.Inference(input);
if (res.Any(z => double.IsNaN(z)))
{
res = new double[] { 1.0, 0, 0, 0, 0, 0, 0 }
}
;
line += string.Join(", ", res.Select(z => z.ToString("0.00", CultureInfo.InvariantCulture)));
sb.AppendLine(line);
}
else
{
sb.AppendLine($"{user.Key}, {string.Join(", ", Enumerable.Range(0, 7).Select(z => "0.00"))}");
}
Console.Write($"\r Remains {count--} ");
}
File.WriteAllText("report_two.csv", sb.ToString());
}
}
private static void ReportOne(List <UserData> userRawData, Dictionary <int, string> juxtaposition)
{
var ruleset = new List <IRule>
{
new SimpleLinearRule(
new double[] { 0.25, 0.01, 0.30, 0.0, 0.001, 0 },
new double[] { 0.80, 0.02, 0.80, 0.8, 0.020, 0.05 },
1),
new SimpleLinearRule(
new double[] { 0.25, 0.5, 0.30, 0.0, 0.05, 0 },
new double[] { 0.60, 0.1, 0.60, 0.6, 0.10, 0.05 },
2),
new SimpleLinearRule(
new double[] { 0.00, 0.00, 0.30, 0.0, 0.50, 0 },
new double[] { 0.40, 0.25, 0.60, 0.3, 0.30, 0.15 },
3),
new SimpleLinearRule(
new double[] { 0.00, 0.00, 0.00, 0.0, 0.10, 0.10 },
new double[] { 0.30, 0.20, 0.30, 0.6, 0.40, 0.50 },
4),
new SimpleLinearRule(
new double[] { 0.00, 0.00, 0.00, 0.00, 0.10, 0.00 },
new double[] { 0.20, 0.20, 0.20, 0.15, 0.50, 0.30 },
5),
new SimpleLinearRule(
new double[] { 0.00, 0.00, 0.00, 0.0, 0.10, 0.00 },
new double[] { 0.90, 0.20, 0.30, 0.4, 0.50, 0.05 },
6)
};
var fis = new ANFIS(ruleset);
var groopy = userRawData.ToLookup(z => z.Id, z => z.GroupId).Where(z => z.Count() < 150);
var sb = new StringBuilder();
sb.AppendLine($"ID, {string.Join(", ", Enumerable.Range(0, 7).Select(z => $"Стадия {z}"))}");
var count = groopy.Count();
foreach (var user in groopy)
{
var codes = user.Where(z => juxtaposition.ContainsKey(z)).Select(z => juxtaposition[z]);
if (codes.Any())
{
var overall = (double)codes.Count();
var line = $"{user.Key}, ";
var input = new double[]
{
codes.Count(z => z.Equals("s6-1a")) / overall,
codes.Count(z => z.Equals("s6-1b")) / overall,
codes.Count(z => z.Equals("s6-1c")) / overall,
codes.Count(z => z.Equals("s6-1d")) / overall,
codes.Count(z => z.Equals("s6-2")) / overall,
codes.Count(z => z.Equals("s6-3")) / overall
};
var res = fis.Inference(input);
if (res.Any(z => double.IsNaN(z)))
{
res = new double[] { 1.0, 0, 0, 0, 0, 0, 0 }
}
;
line += string.Join(", ", res.Select(z => z.ToString("0.00", CultureInfo.InvariantCulture)));
sb.AppendLine(line);
}
else
{
sb.AppendLine($"{user.Key}, {string.Join(", ", Enumerable.Range(0, 7).Select(z => "0.00"))}");
}
Console.Write($"\r Remains {count--} ");
}
File.WriteAllText("report_one.csv", sb.ToString());
}
private static void ReportThree(List <UserData> userRawData, SpecialPublicData[] opposition)
{
var ruleset = new List <IRule>
{
new SimpleLinearRule(
new double[] { 0, 0, 0 },
new double[] { 4, 5, 6 },
1),
new SimpleLinearRule(
new double[] { 2, 1, 1 },
new double[] { 8, 8, 9 },
2),
new SimpleLinearRule(
new double[] { 6, 3, 7 },
new double[] { 11, 11, 11 },
3),
new SimpleLinearRule(
new double[] { 0, 5, 0 },
new double[] { 7, 11, 11 },
4),
new SimpleLinearRule(
new double[] { 2, 6, 5 },
new double[] { 9, 11, 11 },
5),
new SimpleLinearRule(
new double[] { 8, 2, 6 },
new double[] { 11, 9, 11 },
6)
};
var fis = new ANFIS(ruleset);
var groopy = userRawData.ToLookup(z => z.Id, z => z.GroupId).Where(z => z.Count() < 150);
var sb = new StringBuilder();
sb.AppendLine($"ID, {string.Join(", ", Enumerable.Range(0, 7).Select(z => $"Стадия {z}"))}");
var count = groopy.Count();
var dic = opposition.ToDictionary(z => z.Id, z => z);
foreach (var user in groopy)
{
var line = $"{user.Key}, ";
var input = new double[] { 0, 0, 0 };
foreach (var pub in user)
{
if (dic.ContainsKey(pub))
{
input[0] = Math.Max(input[0], dic[pub].s7_1);
input[1] = Math.Max(input[1], dic[pub].s7_2);
input[2] = Math.Max(input[2], dic[pub].s7_3);
}
}
var res = fis.Inference(input);
if (res.Any(z => double.IsNaN(z)))
{
res = new double[] { 1.0, 0, 0, 0, 0, 0, 0 }
}
;
line += string.Join(", ", res.Select(z => z.ToString("0.00", CultureInfo.InvariantCulture)));
sb.AppendLine(line);
Console.Write($"\r Remains {count--} ");
}
File.WriteAllText("report_three.csv", sb.ToString());
}
private static void ReportCommon(List <UserData> userRawData, Dictionary <int, string> juxtaposition, SpecialPublicData[] opposition)
{
var ruleset = new List <IRule>
{
new SimpleLinearRule(
new double[] { 0.25, 0.01, 0.30, 0.0, 0.001, 0.00, 0, -1, 0, 0, 0 },
new double[] { 0.80, 0.02, 0.80, 0.8, 0.020, 0.05, 3, 2, 4, 5, 6 },
1),
new SimpleLinearRule(
new double[] { 0.25, 0.5, 0.30, 0.0, 0.05, 0.00, 1, -1, 2, 1, 1 },
new double[] { 0.60, 0.1, 0.60, 0.6, 0.10, 0.05, 6, 3, 8, 8, 9 },
2),
new SimpleLinearRule(
new double[] { 0.00, 0.00, 0.30, 0.0, 0.50, 0.00, 1, -1, 6, 3, 7 },
new double[] { 0.40, 0.25, 0.60, 0.3, 0.30, 0.15, 11, 4, 11, 11, 11 },
3),
new SimpleLinearRule(
new double[] { 0.00, 0.00, 0.00, 0.0, 0.10, 0.10, 0, 0, 0, 5, 0 },
new double[] { 0.30, 0.20, 0.30, 0.6, 0.40, 0.50, 150, 150, 7, 11, 11 },
4),
new SimpleLinearRule(
new double[] { 0.00, 0.00, 0.00, 0.00, 0.10, 0.00, 2, -1, 2, 6, 5 },
new double[] { 0.20, 0.20, 0.20, 0.15, 0.50, 0.30, 150, 4, 9, 11, 11 },
5),
new SimpleLinearRule(
new double[] { 0.00, 0.00, 0.00, 0.0, 0.10, 0.00, 2, -1, 8, 2, 6 },
new double[] { 0.90, 0.20, 0.30, 0.4, 0.50, 0.05, 21, 3, 11, 9, 11 },
6)
};
var fis = new ANFIS(ruleset);
var groopy = userRawData.ToLookup(z => z.Id, z => z.GroupId).Where(z => z.Count() < 150);
var sb = new StringBuilder();
sb.AppendLine($"ID, {string.Join(", ", Enumerable.Range(0, 7).Select(z => $"Стадия {z}"))}");
var trainset = new StringBuilder();
trainset.AppendLine("ID, %s6-1a, %s6-1b, %s6-1c, %s6-1d, %s6-2, %s6-3, s6-2, s6-3, s7-1, s7-2, s7-3");
var count = groopy.Count();
var dic = opposition.ToDictionary(z => z.Id, z => z);
foreach (var user in groopy)
{
var codes = user.Where(z => juxtaposition.ContainsKey(z)).Select(z => juxtaposition[z]);
if (codes.Any())
{
var overall = (double)codes.Count();
var line = $"{user.Key}, ";
var opp = new double[3];
foreach (var pub in user)
{
if (dic.ContainsKey(pub))
{
opp[0] = Math.Max(opp[0], dic[pub].s7_1);
opp[1] = Math.Max(opp[1], dic[pub].s7_2);
opp[2] = Math.Max(opp[2], dic[pub].s7_3);
}
}
var input = new double[]
{
codes.Count(z => z.Equals("s6-1a")) / overall,
codes.Count(z => z.Equals("s6-1b")) / overall,
codes.Count(z => z.Equals("s6-1c")) / overall,
codes.Count(z => z.Equals("s6-1d")) / overall,
codes.Count(z => z.Equals("s6-2")) / overall,
codes.Count(z => z.Equals("s6-3")) / overall,
codes.Count(z => z.Equals("s6-2")),
codes.Count(z => z.Equals("s6-3")),
opp[0],
opp[1],
opp[2]
};
var res = fis.Inference(input);
if (res.Any(z => double.IsNaN(z)))
{
res = new double[] { 1.0, 0, 0, 0, 0, 0, 0 }
}
;
line += string.Join(", ", res.Select(z => z.ToString("0.00", CultureInfo.InvariantCulture)));
sb.AppendLine(line);
trainset.AppendLine($"{user.Key},{string.Join(", ", input.Select(z => z.ToString("0.0000", CultureInfo.InvariantCulture)))}");
}
else
{
sb.AppendLine($"{user.Key}, {string.Join(", ", Enumerable.Range(0, 7).Select(z => "0.00"))}");
}
Console.Write($"\r Remains {count--} ");
}
File.WriteAllText("report_common.csv", sb.ToString());
File.WriteAllText("origin.csv", trainset.ToString());
}
