TRegResultClass Class

Summary

Calculates additional parameters from multiple linear regression.

Hierarchy
TRegResultClass

Subclasses
None

Description

Calculates additional parameters from multiple linear regression. Use this class to calculate regression coefficients (B) confidence interval, regression coefficients standard deviations and residuals.

How to use TRegResultClass class?

1) Create TRegResultClass: TestClass := TRegResultClass.Create;
2) Use TestClass.B and TestClass.YCalc vectors as parameters in MulLinRegress routine call.
3) Calculate additional parameters by using the CalculateRegResults method.

See Also
TMtxMulLinReg
MulLinRegress
TRegStatsClass

Example 1

The following example was taken from TMtxMulLinReg.Recalc method:

    Uses MtxExpr,Regress,StatTools;

    procedure Example;
    var A,V ; Matrix;
      RegressResult: TRegResultClass;
      RegressStatistics: TRegStatsClass
    begin
      RegressResult := TRegResultClass.Create;
      RegressStatistics := TRegStatsClass.Create;
      try
        // missing ... populate A and RegressResult.B with appropriate data
        // use already created RegressResult.B and RegressResult.YCalc vectors
        MulLinRegress(Y,A,RegressResult.B,true,RegressResult.YCalc,V);
        // do complete regression analysis here with desired confidence equal to 3%
        RegressResult.CalculateRegResults(Y,V,RegressStatistics,false,nil,0.03);
      finally
        RegressResult.Destroy;
        RegressStatistics.Destroy;
      end;
    end;

    #include "MtxVecCpp.h"
    #include "Regress.hpp"
    #include "StatTools.hpp"

    void __fastcall Example()
    {
      Matrix A,V;
      TRegResultClass regres = new TRegResultClass(null);
      TRegStatsClass regstat = new TRegStatsClass(null);
      try
      {
        // missing ... populate A and RegressResult.B with appropriate data
        // use already created RegressResult.B and RegressResult.YCalc vectors
        MulLinRegress(Y,A,RegressResult->B,true,regres->YCalc,V);
        // do complete regression analysis here with desired confidence equal to 3%
        regres->CalculateRegResults(Y,V,RegressStatistics,false,null,0.03);
      }
      __ finally
      {
        regress->Free();
        regstat->Free();
      }

    using Dew.Math;
    using Dew.Stats;
    using Dew.Stats.Units;

    namespace Dew.Examples
    {
      private void Example()
      {
        Matrix A = new Matrix(0,0);
        Matrix V = new Matrix(0,0);
        StatTools.TRegResultClass regres = new StatTools.TRegResultClass();
        StatTools.TRegStatsClass regstat = new StatTools.TRegStatsClass();
        Regress.MulLinRegress(Y,A,regres.B,true,regres.YCalc,V);
        regres.CalculateRegResults(regres.Y,V,regstatfalse,null,0.03);
      }
    }

Properties

Name	Summary
B	Regression coefficients.
BConfInt	Regression coefficients confidence intervals.
BStdDev	Regression coefficients standard deviations.
Residuals	Regresion equation residuals.
YCalc	Dependent variable predicted values.

Methods

Name	Summary
CalculateRegResults	Calculates additional parameters from multiple linear regression.