DataLab is a compact statistics package aiming at exploratory data analysis. Please visit the DataLab Web site for more information....


TFrmMLR

The class TFrmMLR supports the management of the user interface of the multiple linear regression (MLR). Please note that you cannot create an instance of this class yourself (the class has no public Create method). In order to manipulate the parameters of the MLR you have first to call the function MLRGui. This will display the MLR user interface and create the instance FrmMLR which is then globally available throughout a script.

The class TFrmMLR provides the following properties and methods:

Properties

 

Methods

Hint: If you use the global instance FrmMLR before you have called MLRGui this will result in an error message ("Could not call proc at ....")

Example: The following little program first loads the dataset "boilpts_500.idt", then creates the user interface of the MLR, loads the descriptors and the target variable and finally calculates the regression model after setting some of the model parameters:

program MLRTest;

var
  FVers : integer;

begin
if LoadDlabFile (GetDLabDir(ddWork)+'boilpts_500.idt', FVers, true) = 0 then
  begin
  MLRGui;
  if FrmMLR.LoadDescriptors ([3,10,11,2]) = 0 then
    if FrmMLR.LoadTargetVar (13) = 0 then
      begin
      FrmMLR.AttachedForm := true;
      FrmMLR.Left := 600;
      FrmMLR.Top := 50;
      FrmMLR.Width := 1000;
      FrmMLR.Height := 800;
      FrmMLR.AttributeDisplayMode := 0;
      FrmMLR.ModelThroughOrigin := false;
      FrmMLR.SymbolColor := clRed;
      FrmMLR.UseClassColors := true;
      FrmMLR.SymbolIndex := 50;
      FrmMLR.ResultsTab := 1;
      FrmMLR.SetColors (1, $FFEEFF,clBlue, $EEFFEE, clgreen);
      FrmMLR.Calculate;
      end;
  end;
end.