1. Not finding help here? Sign up for a free 30min tutor trial with Chegg Tutors
    Dismiss Notice
Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Linear Algebra - checking a system for consistence.

  1. Aug 19, 2011 #1
    1. The problem statement, all variables and given/known data
    Determine if the system is consistent. Do not completely solve the system.
    I don't know if there's an easy way to input a matrix. These are the numbers from an augmented matrix:
    2 / 0 / 0 / -4 / -10
    0 / 3 / 3 / 0 / 0
    0 / 0 / 1 / 4 / -1
    -3 / 2 / 3 / 1 / 5

    2. Relevant equations
    Row operations used: Replacement and scaling.

    3. The attempt at a solution
    So I think I did this right, but the problem assigned didn't have a solution in the back of the book.

    My answer is that the system is not consistent. The operations that I performed, in order, are:
    1) Scale (1/2)R1 and (1/3)R2
    2) Replace R4 with SUM (R4 + (3)R1)
    3) Replace R4 with SUM (R4 + (-2)R2)
    4) Replace R4 with SUM (R4 + (-1)R3)
    5) Replace R4 with SUM (R4 + R3)

    This yielded an R4 of:
    0 0 0 0 -9, which cannot be true.
  2. jcsd
  3. Aug 19, 2011 #2
    Apparently the board reduces more than 1 space to 1 space. Maybe I'll go back and put in slashes.
  4. Aug 19, 2011 #3


    Staff: Mentor

    I'm pretty sure you made a mistake. I get a solution for the system, which means that it is consistent. In any case, you are not asked for a solution - just determine whether the system is consistent. One way to do that is to take the determinant of the 4 x 4 matrix that is the left part of your augmented matrix. If the determinant is nonzero, the system is consistent; otherwise the system is inconsistent.
  5. Aug 19, 2011 #4
    Thanks for the quick response, Mark. I'm going to rework it. Am I wrong to think that if a row (x1 - x4 in this case) is all 0's, and the other side (the constant - I guess) is non-zero, that the system is inconsistent?
    In other words, I thought that by rule
    0(x1) + 0(x2) + 0(x3) + 0(x4) = C
    meant that the system is inconsistent.
  6. Aug 19, 2011 #5
    Yes you are correct, that would mean that it is inconsistent.

    Taking the determinant of a 4x4 matrix is a pain. I suggest thinking about what this row tells you for a sec: 0 / 3 / 3 / 0 / 0. It should give you the information needed to make this 4x4 a 3x3.
  7. Aug 19, 2011 #6
    Jon, I haven't learned how to perform an operation like that yet. I obviously notice that x2 = -x3, but don't know what that means yet.

    I reworked the problem and very quickly found where I made a careless mistake, and amend my answer to "The system is consistent."
    Mark, you said you took it to a solution, right? I didn't finish it, but I found x4 = 1. Is this correct?
    Thanks so much for all the help!
  8. Aug 19, 2011 #7
    How did you get the system is consistent without solving it? Have you learned about determinants? If so that is probably the method they want you to use

    All I suggested was realizing x3 = -x2, so you could clean up the system a bit to make it easier on yourself before you throw it in a matrix. Just perform the substitution of x3 = -x2.

    So that means you could deal with this instead:
    2 / 0 / -4 / -10
    0 / -1 / 4 / -1
    -3 / -1 / 1 / 5
    Last edited: Aug 19, 2011
  9. Aug 19, 2011 #8

    Ray Vickson

    User Avatar
    Science Advisor
    Homework Helper

    This last statement is incorrect: whether or not a system with zero coefficient determinant is inconsistent depends on the right-had-side. For some right-hand-sides the system will have a (non-unique) solution, so will be consistent, but for others there will be no solution. Simple example: x+y=1 and 2x=2y=2 are consistent, while x=y=1 and 2x+2y=3 are inconsistent.

  10. Aug 19, 2011 #9
    I should have said that I didn't solve it COMPLETELY. I took it to reduced echelon form and confirmed that there was no row with a pivot in the augmented column, or in other words, of the form [0 0 0 b] where b is nonzero.

    Thanks for the tip about what seems like some sort of column simplification step. Using your roadmap, I'll work out how you simplified the matrix.
  11. Aug 19, 2011 #10


    Staff: Mentor

    What I said about consistency and the determinant can be applied only to systems of n equations in n unknowns, so the the related matrices are square.

    In your first example, the system could be written as
    x + y = 1
    x + 0y= 1
    0x + y = 1

    The system for the second example could be written as
    x + y = 3/2
    x + 0y= 1
    0x + y = 1

    Both systems have three equations in two unknowns, so determinant methods aren't applicable.

    If you have a system of n equations in n unknowns (as in the OP's problem, with n = 4), a nonzero determinant means that the system is consistent and that there is a unique solution.
  12. Aug 19, 2011 #11

    Ray Vickson

    User Avatar
    Science Advisor
    Homework Helper

    Sorry: those were typos; I am using a new, unfamiliar keyboard. I meant: first system is x+y=1 and 2x+2y=2; the second system is x+y=1 and 2x+2y=3. Anyway, there is a theorem in linear algebra that says for an nxn system, the equations Ax=b have a unique solution if det(A) is nonzero, and have either no solution or infinitely many solutions if det(A) = 0. Whether or not there are any solutions in the latter case depends on b.

  13. Aug 19, 2011 #12


    Staff: Mentor

    I thought the first one might be a typo, but the second system had what turns out to be the same typo, so I thought you meant x = y = 1. Anyway, now I get what happened.

    Regarding the LA theorem, that's what I was using when I was talking about determinants.
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Similar Discussions: Linear Algebra - checking a system for consistence.