topic DGELS results do not seem correct. in Intel® oneAPI Math Kernel Library

DGELS results do not seem correct.

cy8z — Fri, 23 Jul 2004 10:30:41 GMT

When I call DGELS using the following data: double dA[] = { 1,4,1,5,1,6,1,7,1,8,1,9,1,10,1,11,1,12,1,13,1,14 };
double dB[] = {3.29907758366425,4.46748916656078,19.4376447934279,1.33605625551027,5.63502536637876,5.81066320575566,10.7498837262887,18.834615880912,18.9891064366877,16.2408547103099,5.6606499534616};
char sMatrixType[] = "N";
intiArows = 11;
intiAcols = 2;
intiBcols = 1;
double dwork[512];
intiinfo = 0;
intiwork = 512;

dgels(sMatrixType, &iArows, &iAcols, &iBcols, dA, &iArows, dB, &iArows, dwork, &iwork, &iinfo); For AX=B,DGELS returns1.5727, 0.7746. However, when I do a Least Squares calculation by hand, I get 2.0509, 0.8878. Should SGELS return a solution w/ error minimized? Please help. Best regards, EP.

Re: DGELS results do not seem correct.

lonsunshine — Wed, 25 Aug 2004 01:36:46 GMT

I think your hand calculation is wrong.

The solution given by DGELS

mecej4 — Tue, 06 Aug 2013 01:26:00 GMT

The solution given by DGELS is correct. Matlab gives:[bash]A=reshape([1,4,1,5,1,6,1,7,1,8,1,9,1,10,1,11,1,12,1,13,1,14],[11,2]);

b=[3.29907758366425,4.46748916656078,19.4376447934279,1.33605625551027,5.63502536637876,...

5.81066320575566,10.7498837262887,18.834615880912,18.9891064366877,16.2408547103099,5.6606499534616]';

x=A\\b

x =

1.5728
0.7747

[/bash]