public ActionResult ViewForm()
{
Models.LearningFormModel form = TempData["form"] as Models.LearningFormModel;
var dirPath = Path.Combine(Server.MapPath("~/App_Data/"), form.folderName);
List <string> files = Directory.GetFiles(dirPath).ToList();
string filePath = Path.Combine(dirPath, files.Find(s => s.Contains(form.fileName)));
ErrorBackpropagation backprop = new ErrorBackpropagation(filePath);
double[][] learningTable;
double[][] testingTable;
Network network;
backprop.testNetworkFromFile(out learningTable, out testingTable, out network);
string layers = "";
foreach (Neuron[] layer in network.layers)
{
layers = string.Concat(layers, layer.Length + " ");
}
ViewBag.LearningTable = learningTable;
ViewBag.TestingTable = testingTable;
ViewBag.Layers = layers;
ViewBag.LearningParameter = network.learningParameter;
ViewBag.ErrorTolerance = network.outputTolerance;
return(View(form));
}
public ActionResult Index(Models.LearningFormModel form)
{
if (form != null)
{
TempData["form"] = form;
return(RedirectToAction("ViewForm"));
}
return(View());
}
public ActionResult ViewForm()
{
Models.LearningFormModel form = TempData["form"] as Models.LearningFormModel;
var dirPath = Path.Combine(Server.MapPath("~/App_Data/"), form.folderName);
List <string> files = Directory.GetFiles(dirPath).ToList();
string filePath = Path.Combine(dirPath, files.Find(s => s.Contains(form.fileName)));
ErrorBackpropagation backprop = new ErrorBackpropagation(form.layers, filePath, form.learningParameter, form.errorTolerance, form.successRate, dirPath);
double[] learningRate;
double[] errorRate;
double[][] learningTable;
double[][] testingTable;
backprop.execute(out learningRate, out errorRate, out learningTable, out testingTable);
if (backprop.outputs == 2)
{
string.Concat(form.layers, " 1");
}
else
{
string.Concat(form.layers, " 2");
}
if (learningRate.Length < 100)
{
ViewBag.LearningRate = learningRate.Select(x => x.ToString()).ToArray();
ViewBag.ErrorRate = errorRate.Select(x => x.ToString()).ToArray();
ViewBag.XAxisData = Enumerable.Range(1, errorRate.Length).ToArray();
}
else
{
Array.Copy(learningRate.Select(x => x.ToString()).ToArray(), learningRate.Length - 101, ViewBag.LearningRate, 0, 100);
Array.Copy(errorRate.Select(x => x.ToString()).ToArray(), errorRate.Length - 101, ViewBag.ErrorRate, 0, 100);
ViewBag.XAxisData = Enumerable.Range(1, 100).ToArray();
}
string chartFileName = "chart.txt";
string chartFilePath = Path.Combine(Server.MapPath("~/App_Data/"), chartFileName);
FileStream stream = new FileStream(chartFilePath, FileMode.OpenOrCreate, FileAccess.Write);
StreamWriter writer = new StreamWriter(stream);
for (int i = 0; i < ViewBag.XAxisData.Length; i++)
{
writer.Write(ViewBag.XAxisData[i] + " ");
}
writer.WriteLine();
for (int i = 0; i < ViewBag.ErrorRate.Length; i++)
{
writer.Write(ViewBag.ErrorRate[i] + " ");
}
writer.WriteLine();
for (int i = 0; i < ViewBag.LearningRate.Length; i++)
{
writer.Write(ViewBag.LearningRate[i] + " ");
}
writer.Close();
stream.Close();
ViewBag.LearningTable = learningTable;
ViewBag.TestingTable = testingTable;
return(View(form));
}
