I have a system of $n+1$ equations with $n$ unknowns for which I know the existence of a unique solution. I am interested in writing out an explicit formula for this solution (I am not concerned with speed of computation). For $n \times n$ systems, one could use Cramer's rule (http://mathworld.wolfram.com/CramersRule.html). Is there an analogous result for $m \times n$ systems with a unique solution?

  • 4
    $\begingroup$ Are these all linear equations? If so, one of them is redundant and you can get rid of it $\endgroup$ – Cocopuffs Aug 24 '13 at 23:53
  • $\begingroup$ crammers rule is useful when case has as many equations as number of unknowns , where you have a formula.Roughly speaking it is for square matrices . In other cases use rank method gaussian elimination to solve . The above is for $m \times n $ system , if you can reduce equations back to $n \times n $ then you can thereafter use crammers rule $\endgroup$ – Harish Kayarohanam Aug 25 '13 at 0:05

Yes. If you know the existence of a unique solution then what you can do is to calculate the determinant of submatrices until you find a nonzero one. In particular, select $n$ of the rows of your $m \times n$ matrix and compute the determinant. If it is nonzero then those rows are linearly independent and those $n$-equations suffice to determine the system. At that point, you can use Cramer's Rule on that sub-system.

All of this said, you could just use $\text{rref[ A|b]}$ and read off the answer just the same.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.