// load and setup an existing model
private void DisplayPage_LoadModel()
{
var page = new WizardPage();
var checks = new List <CheckBox>();
var tboxes = new List <TextBox>();
page.hasBackButton = true;
page.hasCancelButton = true;
page.hasNextButton = true;
page.hasFinishButton = false;
page.cancelAction = () =>
{
page.Close();
};
page.backAction = () =>
{
page.Close();
Show();
};
page.nextAction = () =>
{
if (CurrentModel.ModelSource == ANNModel.ModelSourceType.Embedded)
{
try
{
using (MemoryStream ms = new MemoryStream())
using (FileStream file = new FileStream(CurrentModel.ModelPath, FileMode.Open, FileAccess.Read))
{
byte[] bytes = new byte[file.Length];
file.Read(bytes, 0, (int)file.Length);
ms.Write(bytes, 0, (int)file.Length);
SimObject.Model = CurrentModel;
ms.Position = 0;
SimObject.Model.SerializedModelData = Classes.Utils.StreamToBase64(ms);
MessageBox.Show(String.Format("Model successfully embedded in Unit Operation.", CurrentModel.ModelPath), "Embed Model", MessageBoxType.Information);
}
}
catch (Exception ex)
{
MessageBox.Show(ex.ToString(), "Error", MessageBoxType.Error);
}
}
Application.Instance.Invoke(() => DisplayPage_DefineLabels());
page.Close();
};
page.Title = "Neural Network Model Wizard";
page.HeaderTitle = "Step 2 - Load Model";
page.HeaderDescription = "Load an Existing Model from a ZIP file.";
page.FooterText = "Click 'Next' to continue.";
var dl = c.GetDefaultContainer();
dl.Width = Width;
dl.CreateAndAddLabelRow("Load Model from Zip File");
var filepicker = new FilePicker {
Title = "Load Model from Zip File", FileAction = Eto.FileAction.OpenFile
};
filepicker.Filters.Add(new FileFilter("Zip File", new string[] { ".zip" }));
filepicker.FilePathChanged += (s, e) =>
{
CurrentModel.ModelPath = filepicker.FilePath;
CurrentModel.ModelName = Path.GetFileNameWithoutExtension(CurrentModel.ModelPath);
};
dl.CreateAndAddControlRow(filepicker);
dl.CreateAndAddEmptySpace();
dl.CreateAndAddCheckBoxRow("Embed Model Data", true, (chk, e) =>
{
if (chk.Checked.GetValueOrDefault())
{
CurrentModel.ModelSource = ANNModel.ModelSourceType.Embedded;
}
else
{
CurrentModel.ModelSource = ANNModel.ModelSourceType.FileSystem;
}
});
page.Init(Width, Height);
page.Topmost = false;
var scrollable = new Scrollable {
Content = dl, Border = BorderType.None, Size = new Size(Width, Height), ExpandContentHeight = true
};
page.ContentContainer.Add(scrollable);
page.Show();
}
private void DisplayPage_Results()
{
var page = new WizardPage();
var checks = new List <CheckBox>();
var tboxes = new List <TextBox>();
page.hasBackButton = true;
page.hasCancelButton = true;
page.hasNextButton = true;
page.hasFinishButton = false;
page.cancelAction = () =>
{
page.Close();
};
page.backAction = () =>
{
page.Close();
DisplayPage_ML();
};
page.nextAction = () =>
{
Application.Instance.Invoke(() => DisplayPage_SaveModel());
page.Close();
};
page.Title = "Neural Network Model Wizard";
page.HeaderTitle = "Step 7 - Trained Model Predictions";
page.HeaderDescription = "View and Compare the Results Predicted by the Trained Model";
page.FooterText = "Click 'Next' to Continue.";
var dl = c.GetDefaultContainer();
dl.Width = Width;
var plot = new Eto.OxyPlot.Plot();
plot.Model = new PlotModel();
plot.Model.Background = OxyPlot.OxyColors.White;
plot.Model.TitleFontSize = 12;
plot.Model.SubtitleFontSize = 10;
plot.Model.Axes.Add(new OxyPlot.Axes.LinearAxis()
{
MajorGridlineStyle = OxyPlot.LineStyle.Dash,
MinorGridlineStyle = OxyPlot.LineStyle.Dot,
Position = OxyPlot.Axes.AxisPosition.Bottom,
FontSize = 10,
Title = "Input Value (Scaled)",
Key = "x",
});
plot.Model.Axes.Add(new OxyPlot.Axes.LinearAxis()
{
MajorGridlineStyle = OxyPlot.LineStyle.Dash,
MinorGridlineStyle = OxyPlot.LineStyle.Dot,
Position = OxyPlot.Axes.AxisPosition.Left,
FontSize = 10,
Title = "Predicted Value (Scaled)",
});
plot.Model.LegendFontSize = 11;
plot.Model.LegendPlacement = OxyPlot.LegendPlacement.Outside;
plot.Model.LegendOrientation = OxyPlot.LegendOrientation.Horizontal;
plot.Model.LegendPosition = OxyPlot.LegendPosition.BottomCenter;
plot.Model.TitleHorizontalAlignment = OxyPlot.TitleHorizontalAlignment.CenteredWithinView;
plot.Model.Title = "Training Dataset";
plot.Width = 380;
var plot2 = new Eto.OxyPlot.Plot();
plot2.Model = new PlotModel();
plot2.Model.Background = OxyPlot.OxyColors.White;
plot2.Model.TitleFontSize = 12;
plot2.Model.SubtitleFontSize = 10;
plot2.Model.Axes.Add(new OxyPlot.Axes.LinearAxis()
{
MajorGridlineStyle = OxyPlot.LineStyle.Dash,
MinorGridlineStyle = OxyPlot.LineStyle.Dot,
Position = OxyPlot.Axes.AxisPosition.Bottom,
FontSize = 10,
Title = "Input Value (Scaled)",
Key = "x",
});
plot2.Model.Axes.Add(new OxyPlot.Axes.LinearAxis()
{
MajorGridlineStyle = OxyPlot.LineStyle.Dash,
MinorGridlineStyle = OxyPlot.LineStyle.Dot,
Position = OxyPlot.Axes.AxisPosition.Left,
FontSize = 10,
Title = "Predicted Value (Scaled)"
});
plot2.Model.LegendFontSize = 11;
plot2.Model.LegendPlacement = OxyPlot.LegendPlacement.Outside;
plot2.Model.LegendOrientation = OxyPlot.LegendOrientation.Horizontal;
plot2.Model.LegendPosition = OxyPlot.LegendPosition.BottomCenter;
plot2.Model.TitleHorizontalAlignment = OxyPlot.TitleHorizontalAlignment.CenteredWithinView;
plot2.Model.Title = "Testing Dataset";
plot2.Width = 380;
var minscale = CurrentModel.Parameters.MinScale;
var maxscale = CurrentModel.Parameters.MaxScale;
// training plot
plot.Model.AddLineSeries(new double[] { minscale, maxscale }, new double[] { minscale, maxscale }, OxyColors.Black);
for (var i = 0; i < CurrentModel.y_train_unscaled.shape[1]; i++)
{
var xseries = new List <float>();
for (var j = 0; j < CurrentModel.y_train_unscaled.shape[0]; j++)
{
xseries.Add(CurrentModel.y_train_unscaled[j][i]);
}
var yseries = new List <float>();
for (var j = 0; j < CurrentModel.yp_train_unscaled.shape[0]; j++)
{
yseries.Add(CurrentModel.yp_train_unscaled[j][i]);
}
var min = xseries.Min();
var max = xseries.Max();
var xseries2 = xseries.Select(x => (double)x).ToList();
xseries2 = xseries2.Select(x => Utils.Scale(x, min, max, minscale, maxscale)).ToList();
var yseries2 = yseries.Select(x => (double)x).ToList();
yseries2 = yseries2.Select(y => Utils.Scale(y, min, max, minscale, maxscale)).ToList();
plot.Model.AddScatterSeries(xseries2, yseries2);
plot.Model.Series.Last().Title = CurrentModel.Parameters.Labels_Outputs[i];
((ScatterSeries)plot.Model.Series.Last()).MarkerSize = 3.0;
((ScatterSeries)plot.Model.Series.Last()).MarkerType = MarkerType.Circle;
((ScatterSeries)plot.Model.Series.Last()).Title = CurrentModel.Parameters.Labels_Outputs[i];
}
// testing plot
plot2.Model.AddLineSeries(new double[] { minscale, maxscale }, new double[] { minscale, maxscale }, OxyColors.Black);
for (var i = 0; i < CurrentModel.y_test_unscaled.shape[1]; i++)
{
var xseries = new List <float>();
for (var j = 0; j < CurrentModel.y_test_unscaled.shape[0]; j++)
{
xseries.Add(CurrentModel.y_test_unscaled[j][i]);
}
var yseries = new List <float>();
for (var j = 0; j < CurrentModel.yp_test_unscaled.shape[0]; j++)
{
yseries.Add(CurrentModel.yp_test_unscaled[j][i]);
}
var min = xseries.Min();
var max = xseries.Max();
var xseries2 = xseries.Select(x => (double)x).ToList();
xseries2 = xseries2.Select(x => Utils.Scale(x, min, max, minscale, maxscale)).ToList();
var yseries2 = yseries.Select(x => (double)x).ToList();
yseries2 = yseries2.Select(y => Utils.Scale(y, min, max, minscale, maxscale)).ToList();
plot2.Model.AddScatterSeries(xseries2, yseries2);
plot2.Model.Series.Last().Title = CurrentModel.Parameters.Labels_Outputs[i];
((ScatterSeries)plot2.Model.Series.Last()).MarkerSize = 3.0;
((ScatterSeries)plot2.Model.Series.Last()).MarkerType = MarkerType.Circle;
((ScatterSeries)plot2.Model.Series.Last()).Title = CurrentModel.Parameters.Labels_Outputs[i];
}
plot.Model.InvalidatePlot(true);
plot2.Model.InvalidatePlot(true);
var tl = new TableLayout(new TableRow(plot, plot2))
{
Spacing = new Size(10, 10), Height = 440
};
dl.CreateAndAddControlRow(tl);
page.Init(Width, Height);
page.Topmost = false;
var scrollable = new Scrollable {
Content = dl, Border = BorderType.None, Size = new Size(Width, Height), ExpandContentHeight = true
};
page.ContentContainer.Add(scrollable);
page.Show();
}
private void DisplayPage_SaveModel()
{
var page = new WizardPage();
var checks = new List <CheckBox>();
var tboxes = new List <TextBox>();
page.hasBackButton = true;
page.hasCancelButton = true;
page.hasNextButton = false;
page.hasFinishButton = true;
page.cancelAction = () =>
{
page.Close();
};
page.backAction = () =>
{
page.Close();
DisplayPage_ML();
};
page.finishAction = () =>
{
page.Close();
SimObject.UpdateEditForm();
if (!GlobalSettings.Settings.OldUI)
{
SimObject.FlowSheet.UpdateOpenEditForms();
}
};
page.Title = "Neural Network Model Wizard";
page.HeaderTitle = "Step 8 - Save Model";
page.HeaderDescription = "Save the Model to a File or Embed it on the Unit Operation Block.";
page.FooterText = "Click 'Finish' to close this Wizard.";
var dl = c.GetDefaultContainer();
dl.Width = Width;
dl.CreateAndAddLabelRow("Save Model to Zip File");
var filepicker = new FilePicker {
Title = "Save Model to Zip File", FileAction = Eto.FileAction.SaveFile
};
filepicker.Filters.Add(new FileFilter("Zip File", new string[] { ".zip" }));
filepicker.FilePathChanged += (s, e) => CurrentModel.ModelPath = filepicker.FilePath;
dl.CreateAndAddControlRow(filepicker);
dl.CreateAndAddBoldLabelAndButtonRow("Save Model", "Save", null, (btn, e) =>
{
try
{
if (File.Exists(CurrentModel.ModelPath))
{
MessageBox.Show("The save file already exists. Please create a new file and try again.", "Error", MessageBoxType.Error);
return;
}
Classes.Utils.SaveGraphToZip(CurrentModel.session, CurrentModel, CurrentModel.ModelPath);
SimObject.Model = CurrentModel;
SimObject.InitializeMappings();
MessageBox.Show(String.Format("Model saved successfully to '{0}'", CurrentModel.ModelPath), "Save Model", MessageBoxType.Information);
}
catch (Exception ex)
{
MessageBox.Show(ex.ToString(), "Error", MessageBoxType.Error);
}
});
dl.CreateAndAddLabelRow("Embed Model Data");
dl.CreateAndAddBoldLabelAndButtonRow("Embed Model", "Embed", null, (btn, e) =>
{
try
{
using (var ms = new MemoryStream())
{
Classes.Utils.SaveGraphToZipStream(CurrentModel.session, CurrentModel, ms);
SimObject.Model = CurrentModel;
ms.Position = 0;
SimObject.Model.SerializedModelData = Classes.Utils.StreamToBase64(ms);
SimObject.InitializeMappings();
MessageBox.Show(String.Format("Model successfully embedded in Unit Operation.", CurrentModel.ModelPath), "Embed Model", MessageBoxType.Information);
}
}
catch (Exception ex)
{
MessageBox.Show(ex.ToString(), "Error", MessageBoxType.Error);
}
});
page.Init(Width, Height);
page.Topmost = false;
var scrollable = new Scrollable {
Content = dl, Border = BorderType.None, Size = new Size(Width, Height), ExpandContentHeight = true
};
page.ContentContainer.Add(scrollable);
page.Show();
}
public void Show()
{
var page1 = new WizardPage();
page1.hasBackButton = false;
page1.hasCancelButton = true;
page1.hasNextButton = true;
page1.hasFinishButton = false;
page1.cancelAction = () =>
{
page1.Close();
};
page1.Title = "Neural Network Model Wizard";
page1.HeaderTitle = "Step 1 - Model Management";
page1.HeaderDescription = "Select an Action";
page1.FooterText = "Click 'Next' to Continue.";
var dl = c.GetDefaultContainer();
dl.Width = Width;
dl.CreateAndAddLabelRow2("Select an action:");
dl.CreateAndAddEmptySpace();
dl.CreateAndAddEmptySpace();
var rl = new RadioButtonList {
Orientation = Orientation.Vertical
};
rl.Spacing = new Size(10, 10);
rl.Items.Add("Create and Train a New Model");
rl.Items.Add("Load an Existing Model");
rl.SelectedIndex = 0;
page1.nextAction = () =>
{
switch (rl.SelectedIndex)
{
case 0:
CurrentModel = new ANNModel();
Application.Instance.Invoke(() => DisplayPage_LoadData());
break;
case 1:
CurrentModel = new ANNModel();
Application.Instance.Invoke(() => DisplayPage_LoadModel());
break;
}
page1.Close();
};
dl.CreateAndAddControlRow(rl);
var scrollable = new Scrollable {
Content = dl, Border = BorderType.None, Size = new Size(Width, Height), ExpandContentHeight = true
};
page1.Init(Width, Height);
page1.Topmost = false;
page1.ContentContainer.Add(scrollable);
page1.Show();
}
private void DisplayPage_ML()
{
var page = new WizardPage();
var checks = new List <CheckBox>();
var tboxes = new List <TextBox>();
page.hasBackButton = true;
page.hasCancelButton = true;
page.hasNextButton = true;
page.hasFinishButton = false;
page.cancelAction = () =>
{
page.Close();
};
page.backAction = () =>
{
Application.Instance.Invoke(() => DisplayPage_ModelParameters());
page.Close();
};
page.nextAction = () =>
{
Application.Instance.Invoke(() => DisplayPage_Results());
page.Close();
};
page.Title = "Neural Network Model Wizard";
page.HeaderTitle = "Step 6 - Model Training and Evaluation";
page.HeaderDescription = "Click on the 'Train and Evaluate' Button to Train your Model";
page.FooterText = "Click 'Next' to Continue.";
var dl = c.GetDefaultContainer();
dl.Width = Width;
TextArea tb = new TextArea
{
Width = 350,
Font = new Font(FontFamilies.Monospace, 10.0f)
};
var plot = new Eto.OxyPlot.Plot();
plot.Model = new PlotModel();
plot.Model.Background = OxyPlot.OxyColors.White;
plot.Model.TitleFontSize = 12;
plot.Model.SubtitleFontSize = 10;
plot.Model.Axes.Add(new OxyPlot.Axes.LinearAxis()
{
MajorGridlineStyle = OxyPlot.LineStyle.Dash,
MinorGridlineStyle = OxyPlot.LineStyle.Dot,
Position = OxyPlot.Axes.AxisPosition.Bottom,
FontSize = 10,
Title = "Epoch",
Key = "x",
});
plot.Model.Axes.Add(new OxyPlot.Axes.LinearAxis()
{
MajorGridlineStyle = OxyPlot.LineStyle.Dash,
MinorGridlineStyle = OxyPlot.LineStyle.Dot,
Position = OxyPlot.Axes.AxisPosition.Left,
FontSize = 10,
Title = "MSE"
});
plot.Model.LegendFontSize = 11;
plot.Model.LegendPlacement = OxyPlot.LegendPlacement.Outside;
plot.Model.LegendOrientation = OxyPlot.LegendOrientation.Horizontal;
plot.Model.LegendPosition = OxyPlot.LegendPosition.BottomCenter;
plot.Model.TitleHorizontalAlignment = OxyPlot.TitleHorizontalAlignment.CenteredWithinView;
plot.Model.AddLineSeries(new double[] { }, new double[] { }, OxyColors.Red, "Training");
plot.Model.AddLineSeries(new double[] { }, new double[] { }, OxyColors.Blue, "Testing");
plot.Model.Title = "Model Training Results";
var tl = new TableLayout(new TableRow(tb, plot))
{
Spacing = new Size(10, 10), Height = 440
};
dl.CreateAndAddLabelAndButtonRow("Training Results", "Train and Evaluate", null, (btn, e) =>
{
Task.Factory.StartNew(() =>
{
CurrentModel.PrepareData();
CurrentModel.Train(null, tb, plot);
});
});
dl.CreateAndAddControlRow(tl);
page.Init(Width, Height);
page.Topmost = false;
var scrollable = new Scrollable {
Content = dl, Border = BorderType.None, Size = new Size(Width, Height), ExpandContentHeight = true
};
page.ContentContainer.Add(scrollable);
page.Show();
}
private void DisplayPage_LabelData()
{
var page = new WizardPage();
var checks = new List <CheckBox>();
var tboxes = new List <TextBox>();
page.hasBackButton = true;
page.hasCancelButton = true;
page.hasNextButton = true;
page.hasFinishButton = false;
page.cancelAction = () =>
{
page.Close();
};
page.backAction = () =>
{
Application.Instance.Invoke(() => DisplayPage_DisplayData());
page.Close();
};
page.nextAction = () =>
{
CurrentModel.Parameters.Labels = new List <string>();
CurrentModel.Parameters.Labels_Outputs = new List <string>();
foreach (var item in tboxes)
{
CurrentModel.Parameters.Labels.Add(item.Text);
}
int i1 = 0;
foreach (var item in checks)
{
if (item.Checked.GetValueOrDefault())
{
CurrentModel.Parameters.Labels_Outputs.Add(CurrentModel.Parameters.Labels[i1]);
}
i1 += 1;
}
Application.Instance.Invoke(() => DisplayPage_ModelParameters());
page.Close();
};
page.Title = "Neural Network Model Wizard";
page.HeaderTitle = "Step 4 - Data Labelling";
page.HeaderDescription = "Label each data column and select the output variables.";
page.FooterText = "Click 'Next' to Continue.";
var dl = c.GetDefaultContainer();
dl.Width = Width;
dl.CreateAndAddLabelRow2("The Output Variables must be grouped at the end of the list.");
int i = 1;
foreach (var item in CurrentModel.Parameters.Labels)
{
var stack = new StackLayout
{
Orientation = Orientation.Horizontal,
HorizontalContentAlignment = Eto.Forms.HorizontalAlignment.Left,
VerticalContentAlignment = Eto.Forms.VerticalAlignment.Center,
Padding = new Padding(5),
Spacing = 10
};
var lb = new Label()
{
Text = "Column " + i.ToString() + ":"
};
var tb = new TextBox {
Text = item, Width = 300
};
var check = new CheckBox {
Text = "Output"
};
i += 1;
stack.Items.Add(new StackLayoutItem(lb));
stack.Items.Add(new StackLayoutItem(tb));
stack.Items.Add(new StackLayoutItem(check));
tboxes.Add(tb);
checks.Add(check);
dl.CreateAndAddControlRow(stack);
}
page.Init(Width, Height);
page.Topmost = false;
var scrollable = new Scrollable {
Content = dl, Border = BorderType.None, Size = new Size(Width, Height), ExpandContentHeight = true
};
page.ContentContainer.Add(scrollable);
page.Show();
}
private void DisplayPage_ModelParameters()
{
var page = new WizardPage();
var checks = new List <CheckBox>();
var tboxes = new List <TextBox>();
page.hasBackButton = true;
page.hasCancelButton = true;
page.hasNextButton = true;
page.hasFinishButton = false;
page.cancelAction = () =>
{
page.Close();
};
page.backAction = () =>
{
Application.Instance.Invoke(() => DisplayPage_LabelData());
page.Close();
};
page.nextAction = () =>
{
Application.Instance.Invoke(() => DisplayPage_ML());
page.Close();
};
page.Title = "Neural Network Model Wizard";
page.HeaderTitle = "Step 5 - Model Training Parameters";
page.HeaderDescription = "Configure Model Training Parameters";
page.FooterText = "Click 'Next' to Continue.";
var dl = c.GetDefaultContainer();
dl.Width = Width;
var p = CurrentModel.Parameters;
dl.CreateAndAddTextBoxRow(nf, "Minimum Scaled Value", p.MinScale, (tb, e) =>
{
if (tb.Text.IsValidDouble())
{
p.MinScale = (float)tb.Text.ToDoubleFromCurrent();
}
});
dl.CreateAndAddTextBoxRow(nf, "Maximum Scaled Value", p.MaxScale, (tb, e) =>
{
if (tb.Text.IsValidDouble())
{
p.MinScale = (float)tb.Text.ToDoubleFromCurrent();
}
});
dl.CreateAndAddTextBoxRow("N0", "Number of Training Epochs (Iterations)", p.NumberOfEpochs, (tb, e) =>
{
if (tb.Text.IsValidDouble())
{
p.NumberOfEpochs = (int)tb.Text.ToDoubleFromCurrent();
}
});
dl.CreateAndAddTextBoxRow("N0", "Training Batch Size", p.BatchSize, (tb, e) =>
{
if (tb.Text.IsValidDouble())
{
p.BatchSize = (int)tb.Text.ToDoubleFromCurrent();
}
});
dl.CreateAndAddTextBoxRow("N0", "Number of Network Layers (Min = 2, Max = 4)", p.NumberOfLayers, (tb, e) =>
{
if (tb.Text.IsValidDouble())
{
var val = (int)tb.Text.ToDoubleFromCurrent();
if (val < 2)
{
val = 2;
}
if (val > 4)
{
val = 4;
}
p.NumberOfLayers = val;
}
});
dl.CreateAndAddTextBoxRow("N0", "Number of Neurons on First Layer", p.NumberOfNeuronsOnFirstLayer, (tb, e) =>
{
if (tb.Text.IsValidDouble())
{
p.NumberOfNeuronsOnFirstLayer = (int)tb.Text.ToDoubleFromCurrent();
}
});
dl.CreateAndAddTextBoxRow(nf, "Optimizer Learning Rate", p.LearningRate, (tb, e) =>
{
if (tb.Text.IsValidDouble())
{
p.LearningRate = (float)tb.Text.ToDoubleFromCurrent();
}
});
dl.CreateAndAddTextBoxRow("G6", "Minimum Relative Tolerance for MSE", p.RelativeMSETolerance, (tb, e) =>
{
if (tb.Text.IsValidDouble())
{
p.RelativeMSETolerance = (float)tb.Text.ToDoubleFromCurrent();
}
});
dl.CreateAndAddTextBoxRow("G6", "Minimum Absolute Tolerance for MSE", p.AbsoluteMSETolerance, (tb, e) =>
{
if (tb.Text.IsValidDouble())
{
p.AbsoluteMSETolerance = (float)tb.Text.ToDoubleFromCurrent();
}
});
dl.CreateAndAddTextBoxRow(nf, "Split Factor for Data Training/Testing", p.SplitFactor, (tb, e) =>
{
if (tb.Text.IsValidDouble())
{
p.SplitFactor = (float)tb.Text.ToDoubleFromCurrent();
}
});
page.Init(Width, Height);
page.Topmost = false;
var scrollable = new Scrollable {
Content = dl, Border = BorderType.None, Size = new Size(Width, Height), ExpandContentHeight = true
};
page.ContentContainer.Add(scrollable);
page.Show();
}
private void DisplayPage_DisplayData()
{
var page = new WizardPage();
page.hasBackButton = true;
page.hasCancelButton = true;
page.hasNextButton = true;
page.hasFinishButton = false;
page.cancelAction = () =>
{
page.Close();
};
page.backAction = () =>
{
Application.Instance.Invoke(() => DisplayPage_LoadData());
page.Close();
};
page.Title = "Neural Network Model Wizard";
page.HeaderTitle = "Step 3 - Select Data";
page.HeaderDescription = "Select the Data to Train/Test the ANN Model";
page.FooterText = "Click and Drag over the Cells to Select a Data Range and Click 'Next' to Continue.";
var dl = c.GetDefaultContainer();
dl.Width = Width;
var ReoGridControl = new ReoGridFullControl(false)
{
Size = new Size(Width, Height)
};
ReoGridControl.GridControl.bottomPanel.Visible = false;
var stacks = ReoGridControl.Children.Where(x => x is StackLayout).ToList();
foreach (var stack in stacks)
{
stack.Visible = false;
}
dl.CreateAndAddControlRow(ReoGridControl);
page.nextAction = () =>
{
page.Close();
var sheet = ReoGridControl.GridControl.CurrentWorksheet;
int firstcol, lastcol, firstrow, lastrow;
firstcol = sheet.SelectionRange.Col;
lastcol = sheet.SelectionRange.EndCol;
firstrow = sheet.SelectionRange.Row;
lastrow = sheet.SelectionRange.EndRow;
double d;
bool hasheaders = !Double.TryParse(sheet.Cells[firstrow, firstcol].Data.ToString(), out d);
CurrentModel.Parameters.Labels = new List <string>();
CurrentModel.Data = new List <List <double> >();
if (hasheaders)
{
for (int i = firstcol; i <= lastcol; i++)
{
CurrentModel.Parameters.Labels.Add(sheet.Cells[firstrow, i].Data.ToString());
}
int counter = 0;
for (int i = firstrow + 1; i <= lastrow; i++)
{
CurrentModel.Data.Add(new List <double>());
for (int j = firstcol; j <= lastcol; j++)
{
CurrentModel.Data[counter].Add(sheet.Cells[i, j].Data.ToString().ToDoubleFromCurrent());
}
counter += 1;
}
}
else
{
for (int i = firstcol; i <= lastcol; i++)
{
CurrentModel.Parameters.Labels.Add("Column" + i.ToString());
}
int counter = 0;
for (int i = firstrow; i <= lastrow; i++)
{
CurrentModel.Data.Add(new List <double>());
for (int j = firstcol; j <= lastcol; j++)
{
CurrentModel.Data[counter].Add(sheet.Cells[i, j].Data.ToString().ToDoubleFromCurrent());
}
counter += 1;
}
}
Application.Instance.Invoke(() => DisplayPage_LabelData());
};
page.Init(Width, Height);
page.Topmost = false;
page.ContentContainer.Add(dl);
page.Show();
if (File.Exists(CurrentModel.DataSourcePath))
{
DWSIM.UI.Forms.Forms.LoadingData f = null;
f = new DWSIM.UI.Forms.Forms.LoadingData();
f.loadingtext.Text = "Loading Data, Please Wait...";
f.Title = "Loading Data Source";
f.Show();
Application.Instance.AsyncInvoke(() =>
{
try
{
ReoGridControl.GridControl.Load(CurrentModel.DataSourcePath);
ReoGridControl.GridControl.CurrentWorksheet.SetRows(ReoGridControl.GridControl.CurrentWorksheet.MaxContentRow);
ReoGridControl.GridControl.CurrentWorksheet.SetCols(ReoGridControl.GridControl.CurrentWorksheet.MaxContentCol);
}
catch (Exception ex)
{
MessageBox.Show(ex.ToString(), "Error reading data source", MessageBoxType.Error);
}
finally
{
f.Close();
}
});
}
else
{
MessageBox.Show("You didn't select a data source to open. Please go back and try again.", "Error reading data source", MessageBoxType.Error);
}
}
private void DisplayPage_LoadedModelParameters()
{
var page = new WizardPage();
var checks = new List <CheckBox>();
var tboxes = new List <TextBox>();
page.hasBackButton = true;
page.hasCancelButton = true;
page.hasNextButton = false;
page.hasFinishButton = true;
page.cancelAction = () =>
{
page.Close();
};
page.backAction = () =>
{
Application.Instance.Invoke(() => DisplayPage_LabelData());
page.Close();
};
page.finishAction = () =>
{
SimObject.Model = CurrentModel;
SimObject.InitializeMappings();
SimObject.UpdateEditForm();
if (!GlobalSettings.Settings.OldUI)
{
SimObject.FlowSheet.UpdateOpenEditForms();
}
page.Close();
};
page.Title = "Neural Network Model Wizard";
page.HeaderTitle = "Step 4 - Loaded Model Parameters";
page.HeaderDescription = "Configure Loaded Model Parameters";
page.FooterText = "Click 'Finish' to Close this Wizard.";
var dl = c.GetDefaultContainer();
dl.Width = Width;
var p = CurrentModel.Parameters;
dl.CreateAndAddStringEditorRow("Model Name", CurrentModel.ModelName, (tb, e) =>
{
CurrentModel.ModelName = tb.Text;
});
dl.CreateAndAddTextBoxRow(nf, "Minimum Scaled Value", p.MinScale, (tb, e) =>
{
if (tb.Text.IsValidDouble())
{
p.MinScale = (float)tb.Text.ToDoubleFromCurrent();
}
});
dl.CreateAndAddTextBoxRow(nf, "Maximum Scaled Value", p.MaxScale, (tb, e) =>
{
if (tb.Text.IsValidDouble())
{
p.MinScale = (float)tb.Text.ToDoubleFromCurrent();
}
});
dl.CreateAndAddStringEditorRow("Tensor Name for Input Variables (X)", CurrentModel.Parameters.TensorName_X, (tb, e) =>
{
CurrentModel.Parameters.TensorName_X = tb.Text;
});
dl.CreateAndAddStringEditorRow("Tensor Name for Output Variables (Y)", CurrentModel.Parameters.TensorName_Y, (tb, e) =>
{
CurrentModel.Parameters.TensorName_Y = tb.Text;
});
dl.CreateAndAddStringEditorRow("Tensor Name for Output Layer", CurrentModel.Parameters.TensorName_Output, (tb, e) =>
{
CurrentModel.Parameters.TensorName_Output = tb.Text;
});
page.Init(Width, Height);
page.Topmost = false;
var scrollable = new Scrollable {
Content = dl, Border = BorderType.None, Size = new Size(Width, Height), ExpandContentHeight = true
};
page.ContentContainer.Add(scrollable);
page.Show();
}
// create and train a new model
private void DisplayPage_LoadData()
{
var page2 = new WizardPage();
page2.hasBackButton = true;
page2.hasCancelButton = true;
page2.hasNextButton = true;
page2.hasFinishButton = false;
page2.cancelAction = () =>
{
page2.Close();
};
page2.backAction = () =>
{
Application.Instance.Invoke(() => Show());
page2.Close();
};
page2.nextAction = () =>
{
Application.Instance.Invoke(() => DisplayPage_DisplayData());
page2.Close();
};
page2.Title = "Neural Network Model Wizard";
page2.HeaderTitle = "Step 2 - Load Data";
page2.HeaderDescription = "Load Data for Model Training and Testing";
page2.FooterText = "Click 'Next' to Continue.";
var dl = c.GetDefaultContainer();
dl.Width = Width;
dl.CreateAndAddLabelRow2("Select a Data Source. Supported File Formats: XLSX, CSV.");
var filepicker = new FilePicker {
Title = "Select a Data Source"
};
filepicker.Filters.Add(new FileFilter("All Supported File Formats", new string[] { ".xlsx", ".csv", ".txt" }));
filepicker.Filters.Add(new FileFilter("Microsoft Excel Spreadsheets", new string[] { ".xlsx" }));
filepicker.Filters.Add(new FileFilter("Comma-Separated Text Files", new string[] { ".csv", ".txt" }));
if (File.Exists(CurrentModel.DataSourcePath))
{
filepicker.FilePath = CurrentModel.DataSourcePath;
}
filepicker.FilePathChanged += (s, e) => CurrentModel.DataSourcePath = filepicker.FilePath;
dl.CreateAndAddControlRow(filepicker);
var scrollable = new Scrollable {
Content = dl, Border = BorderType.None, Size = new Size(Width, Height), ExpandContentHeight = true
};
page2.Init(Width, Height);
page2.Topmost = false;
page2.ContentContainer.Add(scrollable);
page2.Show();
}
private void DisplayPage_DefineVariableLimits()
{
var page = new WizardPage();
var checks = new List <CheckBox>();
var tboxes = new List <TextBox>();
page.hasBackButton = true;
page.hasCancelButton = true;
page.hasNextButton = true;
page.hasFinishButton = false;
page.cancelAction = () =>
{
page.Close();
};
page.backAction = () =>
{
page.Close();
Application.Instance.Invoke(() => DisplayPage_LoadModel());
};
TextArea tb = new TextArea
{
Height = 300,
Font = new Font(FontFamilies.Monospace, 10.0f)
};
page.nextAction = () =>
{
try
{
var lines = tb.Text.Split(new string[] { "\r\n", "\n" }, StringSplitOptions.RemoveEmptyEntries);
CurrentModel.Parameters.MinValues = new List <float>();
CurrentModel.Parameters.MaxValues = new List <float>();
for (var i = 0; i < lines.Count(); i++)
{
CurrentModel.Parameters.MinValues.Add((float)lines[i].Split(';')[0].ToDoubleFromCurrent());
CurrentModel.Parameters.MaxValues.Add((float)lines[i].Split(';')[1].ToDoubleFromCurrent());
}
Application.Instance.Invoke(() => DisplayPage_LoadedModelParameters());
}
catch (Exception ex)
{
MessageBox.Show(ex.ToString(), "Error", MessageBoxType.Error);
}
page.Close();
};
page.Title = "Neural Network Model Wizard";
page.HeaderTitle = "Step 3A - Define Variable Limits";
page.HeaderDescription = "Define the minimum and maximum values of the variables according to the data used to train the model.";
page.FooterText = "Click 'Next' to continue.";
var dl = c.GetDefaultContainer();
dl.Width = Width;
dl.CreateAndAddLabelRow2("Enter the variables' minimum and maximum values separated by a semicolon (;), including the output ones, one pair per line.");
dl.CreateAndAddControlRow(tb);
page.Init(Width, Height);
page.Topmost = false;
var scrollable = new Scrollable {
Content = dl, Border = BorderType.None, Size = new Size(Width, Height), ExpandContentHeight = true
};
page.ContentContainer.Add(scrollable);
page.Show();
}
private void DisplayPage_DefineLabels()
{
int no = 1;
bool scaled = false;
var page = new WizardPage();
var checks = new List <CheckBox>();
var tboxes = new List <TextBox>();
page.hasBackButton = true;
page.hasCancelButton = true;
page.hasNextButton = true;
page.hasFinishButton = false;
page.cancelAction = () =>
{
page.Close();
};
page.backAction = () =>
{
page.Close();
Application.Instance.Invoke(() => DisplayPage_LoadModel());
};
TextArea tb = new TextArea
{
Height = 300,
Font = new Font(FontFamilies.Monospace, 10.0f)
};
page.nextAction = () =>
{
try
{
var lines = tb.Text.Split(new string[] { "\r\n", "\n" }, StringSplitOptions.RemoveEmptyEntries);
CurrentModel.Parameters.Labels = new List <string>();
CurrentModel.Parameters.Labels_Outputs = new List <string>();
foreach (var line in lines)
{
CurrentModel.Parameters.Labels.Add(line);
}
for (var i = lines.Count() - no; i < lines.Count(); i++)
{
CurrentModel.Parameters.Labels_Outputs.Add(lines[i]);
}
if (scaled)
{
Application.Instance.Invoke(() => DisplayPage_DefineVariableLimits());
}
else
{
Application.Instance.Invoke(() => DisplayPage_LoadedModelParameters());
}
}
catch (Exception ex)
{
MessageBox.Show(ex.ToString(), "Error", MessageBoxType.Error);
}
page.Close();
};
page.Title = "Neural Network Model Wizard";
page.HeaderTitle = "Step 3 - Define Labels";
page.HeaderDescription = "Define the Labels of the Variables in the Model.";
page.FooterText = "Click 'Next' to continue.";
var dl = c.GetDefaultContainer();
dl.Width = Width;
dl.CreateAndAddLabelRow2("Enter the variable labels (names), one per line. The output variables must be the last on the list.");
dl.CreateAndAddControlRow(tb);
dl.CreateAndAddNumericEditorRow("Number of Output Variables", no, 1.0, 100.0, 0, (ns, e) =>
{
no = (int)ns.Value;
});
dl.CreateAndAddCheckBoxRow("Variables are Scaled", scaled, (chk, e) =>
{
scaled = chk.Checked.GetValueOrDefault();
});
page.Init(Width, Height);
page.Topmost = false;
var scrollable = new Scrollable {
Content = dl, Border = BorderType.None, Size = new Size(Width, Height), ExpandContentHeight = true
};
page.ContentContainer.Add(scrollable);
page.Show();
}
