private static List <double[]> preprocessSignal(List <double[]> sinyal)
{
Console.WriteLine("Bandpass Filter");
var watch = new Stopwatch();
watch.Start();
var jumlahKanal = sinyal.Count();
var jumlahSample = sinyal.First().Length;
var sinyalFilter = new List <double[]>();
for (int i = 0; i < jumlahKanal; i++)
{
sinyalFilter.Add(new double[jumlahSample]);
}
Parallel.For(0, jumlahKanal, i => {
var bandpass = new DesignButterworth(Filter.BandPass, 5, 8, 30, _samplingrate);
bandpass.Input = sinyal[i];
bandpass.iirInitialization();
bandpass.compute();
sinyalFilter[i] = bandpass.Output;
});
watch.Stop();
Console.WriteLine("Selesai Bandpass Filter " + watch.ElapsedMilliseconds + " ms");
return(sinyalFilter);
}
private List <double[]> preprocessSignal(List <double[]> sinyal)
{
var jumlahKanal = sinyal.Count();
var jumlahSample = sinyal.First().Length;
var sinyalFilter = new List <double[]>();
for (int i = 0; i < jumlahKanal; i++)
{
sinyalFilter.Add(new double[jumlahSample]);
}
Parallel.For(0, jumlahKanal, i => {
var bandpass = new DesignButterworth(Filter.BandPass, 5, 8, 30, _samplingrate);
bandpass.Input = sinyal[i];
bandpass.iirInitialization();
bandpass.compute();
sinyalFilter[i] = bandpass.Output;
});
return(sinyalFilter);
}
public TestButterworth()
{
InitializeComponent();
var row = File.ReadLines("D:\\Clouds\\Google Drive\\Dari Dropbox\\FILKOM - IF\\Semester 7\\Skripsi\\MuseCSharpLSL\\MuseCSharpLSL\\bin\\Debug\\RAW_HURUF\\Adri_Berhenti_RAW_HURUF_16_56_20.csv");
y = Array.ConvertAll(row.First().Split(';'), double.Parse);
var bandpass = new DesignButterworth(Filter.BandPass, 5, 8, 30, 256);
bandpass.iirInitialization();
bandpass.Input = y;
bandpass.compute();
yf2 = bandpass.Output;
for (int i = 0; i < y.Length; i++)
{
//chart1.Series[0].Points.AddY(y[i]);
chart1.Series[2].Points.AddY(y[i]);
}
}
static void Main(string[] args)
{
var nama = new string[] { "Ramda", "Nanda", "Veny", "Col", "Ekky", "Adri", "Yoga", "RefriR", "RefriR_Test", "Della", "Della_Test", "Galang", "Galang_Test", "Arif", "Arif_Test", "Dian", "Dian_Test", "Chen", "Chen_Test", "Rizky", "Rizky_Test" };
var arah = new string[] { "Maju", "Mundur", "Berhenti", "Kiri", "Kanan" };
double total = nama.Length * arah.Length * 5;
var now = 0;
foreach (var n in nama)
{
foreach (var a in arah)
{
var folderPath = "Sinyal\\" + n + "\\" + a;
foreach (string file in Directory.EnumerateFiles(folderPath, "*.csv"))
{
string dateTime = DateTime.Now.ToString("HH_mm_ss");
var channels = File.ReadLines(file);
var sinyalFilter = new double[channels.Count()][];
int i = 0;
foreach (var c in channels)
{
var signal = Array.ConvertAll(c.Split(';'), double.Parse);
var bandpass = new DesignButterworth(Filter.BandPass, 5, 8, 30, _samplingrate);
bandpass.Input = signal;
bandpass.iirInitialization();
bandpass.compute();
sinyalFilter[i] = bandpass.Output;
i++;
}
var dft = featureExtraction(sinyalFilter);
var feature = dftToOneRowFeature(dft);
saveFile(sinyalFilter, n, "BPF", a, dateTime);
saveFile(dft, n, "DFT", a, dateTime);
saveFeature(feature, n, a);
now++;
Console.WriteLine("DONE " + n + " " + a + " " + now + "/" + total);
}
}
}
Console.ReadKey();
}
static void Main(string[] args)
{
opendatasets();
Console.WriteLine("Masukkan Samplingrate : ");
_SamplingRate = Convert.ToInt32(Console.ReadLine());
Console.WriteLine("Channels (Contoh Input : 0,1,2,..,N (Lihat Dokumentasi Datasets)) : ");
_ChannelsChoose = Array.ConvertAll(Console.ReadLine().Split(','), int.Parse);
Console.WriteLine("Masukkan Times Triggers Start : ");
_TriggerTimeStart = Convert.ToInt32(Console.ReadLine());
Console.WriteLine("Masukkan Times Triggers Stop : ");
_TriggerTimeStop = Convert.ToInt32(Console.ReadLine());
_EpocSizeTraining = (_TriggerTimeStop - _TriggerTimeStart) * _SamplingRate;
_Artifact = new List <List <bool> >();
_ClassLabel = new List <List <int> >();
_Trigger = new List <List <int> >();
_Data = new List <List <double[]> >();
for (int i = 0; i < _dataPath.Count; i++)
{
ReadArtifactSelection(_artifactPath[i]);
Console.WriteLine("Datasets " + Path.GetFileNameWithoutExtension(_dataPath[i]) + " List Artifact Load Done");
}
for (int i = 0; i < _dataPath.Count(); i++)
{
ReadClassLabels(_classLabelPath[i]);
}
for (int i = 0; i < _dataPath.Count; i++)
{
ReadTrigger(_triggerPath[i]);
Console.WriteLine("Datasets " + Path.GetFileNameWithoutExtension(_dataPath[i]) + " List Trigger Load Done");
}
for (int i = 0; i < _dataPath.Count; i++)
{
ReadData(_dataPath[i]);
Console.WriteLine("Datasets " + Path.GetFileNameWithoutExtension(_dataPath[i]) + " List Data EEG Load Done");
}
int countFile = _Data.Count();
_Subject = new double[countFile][][][];//subject //perStimulus //times //Channels
Parallel.For(0, countFile, i =>
{
var perData = new double[_Trigger[i].Count][][];
Parallel.For(0, _Trigger[i].Count(), j =>
{
var startFrom = _Trigger[i][j] + (_SamplingRate * _TriggerTimeStart);
var endTo = startFrom + (_SamplingRate * (_TriggerTimeStop - _TriggerTimeStart));
var newArrayTemp = new double[endTo - startFrom][];
int l = 0;
for (int k = startFrom; k < endTo; k++)
{
newArrayTemp[l] = _Data[i][k];
l++;
}
var pivotArray = new double[newArrayTemp[0].Length][];
Parallel.For(0, pivotArray.Length, k =>
{
var pivotTimes = new double[newArrayTemp.Length];
Parallel.For(0, pivotTimes.Length, m =>
{
pivotTimes[m] = newArrayTemp[m][k];
});
pivotArray[k] = pivotTimes;
});
perData[j] = pivotArray;
});
_Subject[i] = perData;
});
var ApplyingButterworth = new double[countFile][][][];
Parallel.For(0, ApplyingButterworth.Length, i =>
{
var subjectRow = new double[_Subject[i].Length][][];
Parallel.For(0, subjectRow.Length, j =>
{
var channels = new double[_Subject[i][j].Length][];
Parallel.For(0, channels.Length, k =>
{
var bandpass = new DesignButterworth(Filter.BandPass, 5, 8, 30, _SamplingRate);
bandpass.Input = _Subject[i][j][k];
bandpass.iirInitialization();
bandpass.compute();
channels[k] = bandpass.Output;
});
subjectRow[j] = channels;
});
Console.WriteLine("Subject-" + i + " Butterworth Done");
ApplyingButterworth[i] = subjectRow;
});
var ApplyingDFT = new double[countFile][][][];
Parallel.For(0, ApplyingDFT.Length, i =>
{
var subjectRow = new double[ApplyingButterworth[i].Length][][];
Parallel.For(0, subjectRow.Length, j =>
{
var channels = new double[ApplyingButterworth[i][j].Length][];
Parallel.For(0, channels.Length, k =>
{
var jumlahSample = ApplyingButterworth[i][j][k].Length / 2;
channels[k] = new DiscreteFourierTransform().discreteFourierTransform(ApplyingButterworth[i][j][k]).Take(jumlahSample).ToArray();
});
subjectRow[j] = channels;
});
Console.WriteLine("Subject-" + i + " DFT Done");
ApplyingDFT[i] = subjectRow;
});
var ExtractFeature = new double[countFile][][];
Parallel.For(0, ExtractFeature.Length, i =>
{
var subjectRow = new double[ApplyingDFT[i].Length][];
Parallel.For(0, subjectRow.Length, j =>
{
subjectRow[j] = dftToOneRowFeature(ApplyingDFT[i][j]);
});
Console.WriteLine("Subject-" + i + " Extraction Feature Done");
ExtractFeature[i] = subjectRow;
});
Parallel.For(0, ExtractFeature.Length, i =>
{
for (int j = 0; j < ExtractFeature[i].Length; j++)
{
if (!_Artifact[i][j])
{
saveFeature(ExtractFeature[i][j], Path.GetFileNameWithoutExtension(_dataPath[i]), _ClassLabel[i][j]);
}
}
Console.WriteLine("Subject-" + i + " Save Feature Done");
});
}
private void bgProses_DoWork(object sender, DoWorkEventArgs e)
{
this.Invoke((MethodInvoker) delegate
{
tlpMain.Enabled = false;
btnOpenDatasets.Enabled = false;
});
_SamplingRate = (int)numSamplingRate.Value;
_ChannelsChoose = tbChannels.Text.Split(',').Select(Int32.Parse).ToArray();
_TriggerTimeStart = (int)numTriggerStart.Value;
_TriggerTimeStop = (int)numTriggerStop.Value;
_EpocSizeTraining = (_TriggerTimeStop - _TriggerTimeStart) * _SamplingRate;
_Artifact = new List <List <bool> >();
_ClassLabel = new List <List <int> >();
_Trigger = new List <List <int> >();
_Data = new List <List <double[]> >();
for (int i = 0; i < _dataPath.Count; i++)
{
ReadArtifactSelection(_artifactPath[i]);
updateAktivitas("Datasets " + Path.GetFileNameWithoutExtension(_dataPath[i]) + " List Artifact Load Done");
}
for (int i = 0; i < _dataPath.Count(); i++)
{
ReadClassLabels(_classLabelPath[i]);
}
for (int i = 0; i < _dataPath.Count; i++)
{
ReadTrigger(_triggerPath[i]);
updateAktivitas("Datasets " + Path.GetFileNameWithoutExtension(_dataPath[i]) + " List Trigger Load Done");
}
for (int i = 0; i < _dataPath.Count; i++)
{
ReadData(_dataPath[i]);
updateAktivitas("Datasets " + Path.GetFileNameWithoutExtension(_dataPath[i]) + " List Data EEG Load Done");
}
int countFile = _Data.Count();
_Subject = new double[countFile][][][];//subject //perStimulus //times //Channels
Parallel.For(0, countFile, i =>
{
var perData = new double[_Trigger[i].Count][][];
Parallel.For(0, _Trigger[i].Count(), j =>
{
var startFrom = _Trigger[i][j] + (_SamplingRate * _TriggerTimeStart);
var endTo = startFrom + (_SamplingRate * (_TriggerTimeStop - _TriggerTimeStart));
var newArrayTemp = new double[endTo - startFrom][];
int l = 0;
for (int k = startFrom; k < endTo; k++)
{
newArrayTemp[l] = _Data[i][k];
l++;
}
var pivotArray = new double[newArrayTemp[0].Length][];
Parallel.For(0, pivotArray.Length, k =>
{
var pivotTimes = new double[newArrayTemp.Length];
Parallel.For(0, pivotTimes.Length, m =>
{
pivotTimes[m] = newArrayTemp[m][k];
});
pivotArray[k] = pivotTimes;
});
perData[j] = pivotArray;
});
_Subject[i] = perData;
});
var ApplyingButterworth = new double[countFile][][][];
Parallel.For(0, ApplyingButterworth.Length, i =>
{
var subjectRow = new double[_Subject[i].Length][][];
Parallel.For(0, subjectRow.Length, j =>
{
var channels = new double[_Subject[i][j].Length][];
Parallel.For(0, channels.Length, k =>
{
var bandpass = new DesignButterworth(Filter.BandPass, 5, 8, 30, _SamplingRate);
bandpass.Input = _Subject[i][j][k];
bandpass.iirInitialization();
bandpass.compute();
channels[k] = bandpass.Output;
});
subjectRow[j] = channels;
});
updateAktivitas("Subject-" + i + " Butterworth Done");
ApplyingButterworth[i] = subjectRow;
});
var ApplyingDFT = new double[countFile][][][];
Parallel.For(0, ApplyingDFT.Length, i =>
{
var subjectRow = new double[ApplyingButterworth[i].Length][][];
Parallel.For(0, subjectRow.Length, j =>
{
var channels = new double[ApplyingButterworth[i][j].Length][];
Parallel.For(0, channels.Length, k =>
{
var jumlahSample = ApplyingButterworth[i][j][k].Length / 2;
channels[k] = new DiscreteFourierTransform().discreteFourierTransform(ApplyingButterworth[i][j][k]).Take(jumlahSample).ToArray();
});
subjectRow[j] = channels;
});
updateAktivitas("Subject-" + i + " DFT Done");
ApplyingDFT[i] = subjectRow;
});
var ExtractFeature = new double[countFile][][];
Parallel.For(0, ExtractFeature.Length, i =>
{
var subjectRow = new double[ApplyingDFT[i].Length][];
Parallel.For(0, subjectRow.Length, j =>
{
subjectRow[j] = dftToOneRowFeature(ApplyingDFT[i][j]);
});
updateAktivitas("Subject-" + i + " Extraction Feature Done");
ExtractFeature[i] = subjectRow;
});
Parallel.For(0, ExtractFeature.Length, i =>
{
for (int j = 0; j < ExtractFeature[i].Length; j++)
{
if (!_Artifact[i][j])
{
saveFeature(ExtractFeature[i][j], Path.GetFileNameWithoutExtension(_dataPath[i]), _ClassLabel[i][j]);
}
}
updateAktivitas("Subject-" + i + " Save Feature Done");
});
this.Invoke((MethodInvoker) delegate
{
tlpMain.Enabled = true;
btnOpenDatasets.Enabled = true;
});
if (bgProses.CancellationPending)
{
e.Cancel = true;
return;
}
}
static void Main(string[] args)
{
Stopwatch stopwatch = new Stopwatch();
stopwatch.Start();
_Artifact = new List <List <bool> >();
_ClassLabel = new List <List <int> >();
_Trigger = new List <List <int> >();
foreach (string file in Directory.EnumerateFiles("datasets", "*.HDR_ArtifactSelection"))
{
var contents = File.ReadLines(file);
var newArtifact = new List <bool>();
foreach (var row in contents)
{
newArtifact.Add(Convert.ToBoolean(Convert.ToInt16(row)));
}
_Artifact.Add(newArtifact);
}
foreach (string file in Directory.EnumerateFiles("datasets", "*.HDR_Classlabel"))
{
var contents = File.ReadLines(file);
var newLabel = new List <int>();
foreach (var row in contents)
{
if (row == "NaN")
{
newLabel.Add(5);
}
else
{
newLabel.Add(Convert.ToInt16(row));
}
}
_ClassLabel.Add(newLabel);
}
foreach (string file in Directory.EnumerateFiles("datasets", "*.HDR_TRIG"))
{
var contents = File.ReadLines(file);
var newTrigger = new List <int>();
foreach (var row in contents)
{
newTrigger.Add(Convert.ToInt32(row));
}
_Trigger.Add(newTrigger);
}
_Data = new List <List <double[]> >();
foreach (string file in Directory.EnumerateFiles("datasets", "*.data"))
{
var contents = File.ReadLines(file);
var newData = new List <double[]>();
foreach (var row in contents)
{
var timesAllChannels = Array.ConvertAll(row.Split(" ".ToCharArray(), StringSplitOptions.RemoveEmptyEntries), Double.Parse);
var times = new double[] { timesAllChannels[26], timesAllChannels[5], timesAllChannels[9], timesAllChannels[36] };
newData.Add(times);
}
_Data.Add(newData);
}
int countFile = _Data.Count();
_Subject = new double[countFile][][][];//subject //perStimulus //times //Channels
Parallel.For(0, countFile, i =>
{
var perData = new double[_Trigger[i].Count][][];
Parallel.For(0, _Trigger[i].Count(), j =>
{
var startFrom = _Trigger[i][j] + (_SamplingRate * (_FixationCross + _BlankScreen));
var endTo = startFrom + (_SamplingRate * _StimulusShown);
var newArrayTemp = new double[endTo - startFrom][];
int l = 0;
for (int k = startFrom; k < endTo; k++)
{
newArrayTemp[l] = _Data[i][k];
l++;
}
var pivotArray = new double[newArrayTemp[0].Length][];
Parallel.For(0, pivotArray.Length, k =>
{
var pivotTimes = new double[newArrayTemp.Length];
Parallel.For(0, pivotTimes.Length, m =>
{
pivotTimes[m] = newArrayTemp[m][k];
});
pivotArray[k] = pivotTimes;
});
perData[j] = pivotArray;
});
_Subject[i] = perData;
});
var ApplyingButterworth = new double[countFile][][][];
Parallel.For(0, ApplyingButterworth.Length, i =>
{
var subjectRow = new double[_Subject[i].Length][][];
Parallel.For(0, subjectRow.Length, j =>
{
var channels = new double[_Subject[i][j].Length][];
Parallel.For(0, channels.Length, k =>
{
var bandpass = new DesignButterworth(Filter.BandPass, 5, 8, 30, _SamplingRate);
bandpass.Input = _Subject[i][j][k];
bandpass.iirInitialization();
bandpass.compute();
channels[k] = bandpass.Output;
});
subjectRow[j] = channels;
});
Console.WriteLine("Subject-" + i + " Butterworth Done");
ApplyingButterworth[i] = subjectRow;
});
var ApplyingDFT = new double[countFile][][][];
Parallel.For(0, ApplyingDFT.Length, i =>
{
var subjectRow = new double[ApplyingButterworth[i].Length][][];
Parallel.For(0, subjectRow.Length, j =>
{
var channels = new double[ApplyingButterworth[i][j].Length][];
Parallel.For(0, channels.Length, k =>
{
var jumlahSample = ApplyingButterworth[i][j][k].Length / 2;
channels[k] = new DiscreteFourierTransform().discreteFourierTransform(ApplyingButterworth[i][j][k]).Take(jumlahSample).ToArray();
});
subjectRow[j] = channels;
});
Console.WriteLine("Subject-" + i + " DFT Done");
ApplyingDFT[i] = subjectRow;
});
var ExtractFeature = new double[countFile][][];
Parallel.For(0, ExtractFeature.Length, i =>
{
var subjectRow = new double[ApplyingDFT[i].Length][];
Parallel.For(0, subjectRow.Length, j =>
{
subjectRow[j] = dftToOneRowFeature(ApplyingDFT[i][j]);
});
Console.WriteLine("Subject-" + i + " Extraction Feature Done");
ExtractFeature[i] = subjectRow;
});
Parallel.For(0, ExtractFeature.Length, i =>
{
for (int j = 0; j < ExtractFeature[i].Length; j++)
{
if (!_Artifact[i][j])
{
saveFeature(ExtractFeature[i][j], i, _ClassLabel[i][j] - 1);
}
}
Console.WriteLine("Subject-" + i + " Save Feature Done");
});
Console.WriteLine(stopwatch.ElapsedMilliseconds);
Console.ReadKey();
stopwatch.Stop();
}
