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!

Basic Linear Algebra Proof

  1. Jan 21, 2008 #1
    1. The problem statement, all variables and given/known data
    Consider the homogeneous system of linear equations
    [tex] ax + by = 0 [/tex]
    [tex] cx + dy = 0 [/tex]
    Prove that if [tex] ad - bc \not= 0 [/tex], then [tex] x = 0,y=0 [/tex] is the only solution to the system.




    3. The attempt at a solution

    First, I tried rewriting the system of equations to get [tex] y = -\frac{a}{b} x [/tex] and [tex] y = -\frac{c}{d} x [/tex]. This would have probably helped me in the proof, but I realized that I may not be able to divide by [tex] b [/tex] and [tex] d [/tex] because they may be [tex] 0[/tex].

    Maybe I could use the contrapositive to prove this. Proving the statement "If [tex] ad - bc = 0[/tex], then [tex] x = 0, y = 0 [/tex] is not the only solution to the system. I'd have to show that there are more solutions. I am not sure how to do this though. Although, this seems like the easiest way to prove it.



    Can anyone give me some help? Thanks in advance.
     
  2. jcsd
  3. Jan 21, 2008 #2
    Are you familiar with determinants yet?
     
  4. Jan 21, 2008 #3
    No, this is the beginning of the Linear course. We didn't do much more than an introduction to matrices and Reduced Echelon Form for Gauss-Jordan Elimination.

    Oh i'm sorry, I posted the question incorrectly. It was an if and only if proof, and I did that direction pretty easily.

    The implication I actually need help on is the opposite direction.

    If x = 0, y = 0 is the only solution to the system, then [tex] ad - bc \not= 0 [/tex].
     
  5. Jan 21, 2008 #4
    As you said if ad = bc, then you should be able to show that (0,0) is not the only solution. Look at what happens in the original system if you assume this, and you then multiply the top equation by d and the bottom equation by b.
     
  6. Jan 21, 2008 #5

    HallsofIvy

    User Avatar
    Staff Emeritus
    Science Advisor

    Then say "if b is not 0" and "if d is not 0" and see what you get. Then go back and say "if b= 0" or "if d= 0" and see what you get.

    No, that's not the contrapositive. You want to prove "If ad-bc is not 0 then x= 0, y= 0 is the only solution". The contrapositive of that is "if x= 0, y= 0 is not the only solution then ad- bc= 0." I think you would find that harder. Have you considered just going ahead and solving the system? Try starting with ax+ by= 0 and multiplying both sides by d: adx+ bdy= 0. Now take cx+ dy= 0 and multiply both sides by -b: -bcx- bdy= 0. Adding those two equations will eliminate y. Do you see why you need "ad- bc not equal to 0"? What would happen if ad- bc= 0?
     
  7. Jan 21, 2008 #6
    Mathdope:

    assume ad = bc.

    multiplying the top equation by d and the bottom by b:
    [tex]d(ax + by) = 0 [/tex]
    [tex]b(cx + dy) = 0 [/tex]

    [tex] adx + bdy = 0 [/tex]
    [tex] bcx + bdy = 0 [/tex]

    Now i'm sure there's something I can do with ad and bc being the coefficients of x.. since they're equal.

    I'm just not sure though.
     
  8. Jan 21, 2008 #7
    If ad = bc, then adx + bdy = 0 and bcx + bdy = 0 are the same equation. What does that say about the solutions to the system?
     
  9. Jan 21, 2008 #8
    The lines coincide so there are infinitely-many solutions.

    So that's all I need to do for that proof?
     
  10. Jan 21, 2008 #9
    [tex] adx + bdy = 0 [/tex]
    [tex] bcx + bdy = 0 [/tex]

    Notice that these two imply that

    [tex] adx - bcx = 0 [/tex]

    Edit: notice that you don't need to assume that ad = bc to perform the multiplications. Just do them and then see what happens IF ad = bc.
     
    Last edited: Jan 21, 2008
  11. Jan 21, 2008 #10
    Oh, sorry, I didn't read carefully. You either need to show that if there's a solution other than (0,0) then ad = bc or follow what Mathdope suggested to the logical conclusion.
     
  12. Jan 21, 2008 #11
    Mathdope:

    Okay.. so if i factor out an x:

    (ad - bc)x = 0

    since ad = bc,
    we get 0*x = 0.

    But that doesn't tell us anything, does it?

    I'm still confused :(
     
  13. Jan 21, 2008 #12
    If ad = bc, then x can be anything. If ad - bc is nonzero, what must x be? Reread the original problem.
     
    Last edited: Jan 21, 2008
  14. Jan 21, 2008 #13
    the original problem is... "If x = 0, y = 0 is the only solution, then [tex] ad - bc \not= 0 [/tex]."


    To answer your question:

    If ad - bc is nonzero then x must be zero.

    I didn't think I could assume that ad - bc is nonzero in this proof though.

    Since i'm assuming ad - bc = 0... Or similarly, i can assume that x = 0, y = 0 is the only solution.
     
  15. Jan 21, 2008 #14
    If ad - bc = 0 then x can have any value. This means that (x,y) = (0,0) is not the only solution. If you think of x as a parameter t, then any ordered pair of the form (t,-(a/b)t) is a solution (for nonzero b). If b = 0, then (0,t) is a solution for all t. (Notice that if ad = bc, then a/b = c/d when b and d aren't zero).
     
    Last edited: Jan 21, 2008
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?



Similar Discussions: Basic Linear Algebra Proof
Loading...