public void Apply()
{
item.Name = name.Text;
item.Dimensions = (int)Dimensions.Value;
item.ErrorColumns = Errors.Checked;
item.Compile();
Reset();
Model.Items.Update(item);
}
public FitDemo(PlotControl plot)
{
this.plot = plot;
plot.Model.Clear();
// create a new DataItem
data = new DataItem();
data.Dimensions = 2; // This property is only used by the LoadText method, it is not required for the fit algorithm.
data.ErrorColumns = true; // This property is only used by the LoadText method, it is not required for the fit algorithm.
// preset the x, dx and dy values to functions:
data.x.Source = "n/Length*12 - 8"; // place the data in the visible plot range (x0: -8; x1: 4 in this demo application)
data.dx.Source = "0.5"; // set the x errors to 0.5 (the x errors are not used by the fit algorithm)
data.dy.Source = "y[n]*0.2"; // set y errors to 20% of the x values
data.Compile();
data.LoadText("Gauss data.txt", "\n"); // Load the data from the "Gauss data.txt" text file.
// Display data
plot.Model.Add(data);
// create a new function
f = new Function1DItem(@"
// M - gauss curves with derivative information in dfdp.
const int M = 3;
double arg, ex, fac, sum = 0;
for (int n = 0; n < 3*M; n += 3) {
arg = (x - p[n + 1])/p[n + 2];
ex = Math.Exp(-arg*arg);
fac = p[n]*ex*2*arg;
// Compute derivative information in order to speed up the Marquardt fitting algorithm. Marquardt also works with no
// derivative information provided by the funciton, so you can also omit this, it then computes numerical derivatives.
// The NelderMead fitting algorithm uses no derivatives.
dfdp[n] = ex;
dfdp[n + 1] = fac/p[n + 2];
dfdp[n + 2] = fac*arg/p[n + 2];
// compute the sum over all gauss curves.
sum += p[n]*ex;
}
return sum;
" );
f.Compile();
// Set initial fit guess.
ResetGuess();
// Display function
plot.Model.Add(f);
}
private void Compile()
{
data.Compile(true);
if (!data.compiled)
{
string msg = "There are errors in the formulas:\n";
for (int i = 0; i < data.errors.Length; i++)
{
msg += data.errors[i] + "\n";
}
MessageBox.Show(msg, "Compile error");
}
}
public void Apply()
{
item.loadsource = source.Text;
item.Compile(true);
if (!item.compiled)
{
string text = "Compile Errors:\n";
for (int i = 0; i < item.errors.Length; i++)
{
text += item.errors[i] + "\n";
}
MessageBox.Show(text);
}
}
public void Apply()
{
item.name = name.Text;
item.Color = colorLabel.BackColor;
item.lineStyle = (DashStyle)lineStyle.SelectedIndex;
try {
item.lineWidth = float.Parse(lineWidth.Text);
} catch { item.lineWidth = 1; }
item.lines = line.Checked;
item.marks = marks.Checked;
item.Compile(true);
Reset();
if (!deleted)
{
graph.Invalidate();
}
mainform.ResetMenu();
}
private void asciiClick(object sender, System.EventArgs e)
{
DataItem data = new DataItem();
//The loadsource represents the body of the following function:
//void Load(System.IO.FileStream stream) {
// ...
//}
//The function loads the data from the stream and sets the arrays x, y, dx and dy accordingly.
//The size of the arrays is automatically adjusted upon access.
data.loadsource =
"using (StreamReader r = new StreamReader(stream)) {" +
" int n = 0; string line; string[] tokens;" +
" char[] separator = \";,|\".ToCharArray();" +
" while ((line = r.ReadLine()) != null) {" +
" tokens = line.Split(separator);" +
" try {x[n] = double.Parse(tokens[0]);} catch {x[n] = 0;}" +
" try {y[n] = double.Parse(tokens[1]);} catch {y[n] = 0;}" +
" try {dx[n] = double.Parse(tokens[2]);} catch {dx[n] = 0;}" +
" try {dy[n] = double.Parse(tokens[3]);} catch {dy[n] = 0;}" +
" n++;" +
" }" +
"}";
data.Compile(true);
if (!data.compiled)
{
MessageBox.Show("Error in sourcecode:\n" + data.errors[0]);
}
try {
data.LoadFromFile("data.csv");
} catch (Exception ex) {
MessageBox.Show("Could not open the file data.csv\n" + ex.Message);
}
graph.Model.Items.Add(data);
Console.WriteLine("data Length: {0}", data.Length);
Console.WriteLine("x y dx dy");
for (int n = 0; n < data.Length; n++)
{
Console.WriteLine("{0}, {1}, {2}, {3}", data.x[n], data.y[n], data.dx[n], data.dy[n]);
}
graph.Invalidate();
}
