ISmBinaryModel.GuessingTable

Syntax

GuessingTable: ISlQualityTable;

Description

The GuessingTable property returns a collection of values that describe the quality of binary model selection.

Comments

The collection of values forms the following table:

  Fact
(ISlQualitySet.Actual)
Predicted
(ISlQualitySet.Prediction)
Predicted correctly
(ISlQualitySet.CorrectPrediction)
Number: 0 3 3 2
Number: 1 2 2 1

Example

To execute the example, add a link to the Stat system assembly.

Sub UserProc;
Var
    bm: SmBinaryModel;
    can: Array[9Of Double;
    bin2: Array[5Of Integer;
    i, res, mm: Integer;
    Intercept: IIntercept;
    Explanatories: ISlSeries;
    GuessingTable: ISlQualityTable;
    GuessTableItem: ISlQualitySet;
Begin
    bm := New SmBinaryModel.Create;
    // Set values of explanatory series
    can[00] := 6.209; can[05] := 5;
    can[01] := 6.385; can[06] := 6;
    can[02] := 6.29; can[07] := 7;
    can[03] := 6.25; can[08] := 8;
    can[04] := 6.1;
    // Set explained series values
    bin2[00] := 1; bin2[03] := 0;
    bin2[01] := 1; bin2[04] := 0;
    bin2[02] := 0;
   // Set values for the first and the last points of the identification period
    bm.ModelPeriod.FirstPoint := 1;
    bm.ModelPeriod.LastPoint := 5;
    // Set value for the last forecast point
    bm.Forecast.LastPoint := 9;
    // Set model type
    bm.BinaryDistr := BinaryDistrType.Probit;
    // Set value of group division and accuracy
    bm.ClassificationCutOff := 0.5;
    bm.Tolerance := 0.001;  
    // Set method of calculating the constant
    Intercept := bm.ModelCoefficients.Intercept;
    Intercept.Mode := InterceptMode.AutoEstimate;   
    // Set explained series
    bm.BinaryExplained := bin2;
    // Set explanatory series
    Explanatories := bm.Explanatories;
    Explanatories.Add.Value := can;
    // Perform calculation and display error  messages
    res:= bm.Execute;
    // Display calculation results
    If (res = 0Then
        // Table of binary model fitting quality
        Debug.WriteLine("Fitting quality description");
        GuessingTable := bm.GuessingTable;
        For mm := 0 To 1 Do
            GuessTableItem := GuessingTable.Item(mm);
            i := GuessTableItem.Actual;
            Debug.WriteLine(mm.ToString + ". Fact: " + i.ToString);
            i := GuessTableItem.Prediction;
            Debug.WriteLine(mm.ToString + ". Predicted: " + i.ToString);
            i := GuessTableItem.CorrectPrediction;
            Debug.WriteLine(mm.ToString + ". Predicted correctly: " + i.ToString);    
        End For;
        Else
            Debug.WriteLine(bm.Errors);
    End If
End Sub UserProc;

After executing the example the console window displays calculation results and values describing quality of binary model selection:

Selection quality description

0. Fact: 3

0. Predicted: 3

0. Predicted correctly: 2

1. Fact: 2

1. Predicted: 2

1. Predicted correctly: 1

See also:

ISmBinaryModel