With Dew Stats for .NET it's not difficult to do complex tasks. Doing multiple linear regression is a peace of cake:

• Add Dew.Math.Units, Dew.Math.Editors to the using section.
• Drop tMtxMulLinReg component and the Button on the form.
• Bring up the TMtxMulLinReg design time editor by double clicking on MtxMulLinReg1.
• Define if constant term or weights will be used (this can also be done at runtime via code). Close TMtxMulLinReg editor.
• Double click on Button1 and add the following code:

tMtxMulLinReg1.Y.LoadFromFile("c:\temp\yvals.vec");
tMtxMulLinReg1.A.LoadFromFile("c:\temp\system.mtx");
tMtxMulLinReg1.Recalculate();
MtxVecEdit.ViewValues(tMtxMulLinReg1.RegressResult.B,"Regression coefficients",true,false,false);

Other things to try:

• Access RegressResult properties (YCalc, Residuals). Draw these values by using Dew.Math.DrawValues method
• Load Y,A data from ASCII file (demonstrated in demo). Repeat procedure.