private void Bw_RunWorkerCompleted(object sender, RunWorkerCompletedEventArgs e)
{
LottoWorker lw = (LottoWorker)e.Result;
var loadedModel = lw.Model;
var mlContext = new MLContext();
int lastRecord = data.Count - 1;
var predictionEngine = mlContext.Model.CreatePredictionEngine <LottoData, LottoDataPrediction>(loadedModel);
var lt = new LottoData()
{
prev1 = data[lastRecord - 8][lw.Slot - 1],
prev2 = data[lastRecord - 7][lw.Slot - 1],
prev3 = data[lastRecord - 6][lw.Slot - 1],
prev4 = data[lastRecord - 5][lw.Slot - 1],
prev5 = data[lastRecord - 4][lw.Slot - 1],
prev6 = data[lastRecord - 3][lw.Slot - 1],
prev7 = data[lastRecord - 2][lw.Slot - 1],
prev8 = data[lastRecord - 1][lw.Slot - 1],
prev9 = data[lastRecord][lw.Slot - 1]
};
var predict = predictionEngine.Predict(lt);
log("PREDICTED SLOT [" + lw.Slot + "] - " + predict.predictedresult.ToString());
predictedResult.Add(predict.predictedresult.ToString());
if (File.Exists(lw.Filename))
{
File.Delete(lw.Filename);
}
}
private void Bw_DoWork(object sender, DoWorkEventArgs e)
{
LottoWorker lw = (LottoWorker)e.Argument;
lw.Worker.ReportProgress(0, "SLOT [" + lw.Slot + "] - Reading Training Data");
var mlContext = new MLContext();
var loader = mlContext.Data.CreateTextLoader(new[]
{
new TextLoader.Column("prev1", DataKind.Single, 0),
new TextLoader.Column("prev2", DataKind.Single, 1),
new TextLoader.Column("prev3", DataKind.Single, 2),
new TextLoader.Column("prev4", DataKind.Single, 3),
new TextLoader.Column("prev5", DataKind.Single, 4),
new TextLoader.Column("prev6", DataKind.Single, 5),
new TextLoader.Column("prev7", DataKind.Single, 6),
new TextLoader.Column("prev8", DataKind.Single, 7),
new TextLoader.Column("prev9", DataKind.Single, 8),
new TextLoader.Column("result", DataKind.Single, 9),
}, hasHeader: true, separatorChar: ',');
var trainingData = loader.Load(lw.Filename);
var learningPipeline = mlContext.Transforms.Conversion.MapValueToKey("result")
.Append(mlContext.Transforms.Concatenate("Features", "prev1", "prev2", "prev3", "prev4", "prev5", "prev6", "prev7", "prev8", "prev9"))
.Append(mlContext.MulticlassClassification.Trainers.SdcaNonCalibrated(labelColumnName: "result", featureColumnName: "Features"))
.Append(mlContext.Transforms.Conversion.MapKeyToValue("PredictedLabel"));
lw.Worker.ReportProgress(0, "SLOT [" + lw.Slot + "] - Training Data. This may take a lot of time...");
lw.Model = learningPipeline.Fit(trainingData);
e.Result = lw;
}
private void BtnGenerateModel_Click(object sender, EventArgs e)
{
if (data != null && data.Count > 0)
{
groupBox1.Enabled = false;
predictedResult = new List <string>();
List <LottoWorker> workers = new List <LottoWorker>();
var fn = GenerateCSVForTraining(data);
for (int i = 0; i < fn.Length; i++)
{
string f = fn[i];
BackgroundWorker bw = new BackgroundWorker();
bw.WorkerReportsProgress = true;
bw.DoWork += Bw_DoWork;
bw.ProgressChanged += Bw_ProgressChanged;
bw.RunWorkerCompleted += Bw_RunWorkerCompleted;
LottoWorker lw = new LottoWorker();
lw.Worker = bw;
lw.Filename = f;
lw.Slot = i + 1;
workers.Add(lw);
lw.Worker.RunWorkerAsync(lw);
}
while (true)
{
RESTART:
Thread.Sleep(2000);
Application.DoEvents();
foreach (var lw in workers)
{
if (lw.Worker.IsBusy)
{
goto RESTART;
}
}
break;
}
string res = "PREDICTED RESULTS : " + String.Join(" - ", predictedResult.ToArray());
log(res);
MessageBox.Show(res, "Result", MessageBoxButtons.OK, MessageBoxIcon.Information);
groupBox1.Enabled = true;
}
else
{
MessageBox.Show("Please load a dataset via CSV.", "No Data Available", MessageBoxButtons.OK, MessageBoxIcon.Stop);
}
}
