Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Why is <v,v> >= 0 ?

  1. Apr 27, 2012 #1


    User Avatar

    I think I am missing a subtle point of the definition of a inner product. All the texts I have seen state <v,v> >= 0

    If you have say:


    then <v,v> = -3 (Using the definition where you do the dot product, while conjugating the first term)

    This is a negative number and defies the above definition of an inner product.

    Is it that (1,2i) is not in an inner product space and therefore doen't have an inner product?
  2. jcsd
  3. Apr 27, 2012 #2


    User Avatar
    Science Advisor
    Homework Helper

    welcome to pf!

    hi 9k9! welcome to pf! :smile:

    (1,2i).(1,-2i) = 1.1 + 2i.-2i = 1 - 4i2 = 5 :wink:
  4. Apr 28, 2012 #3

    I like Serena

    User Avatar
    Homework Helper

    Welcome to PF, 9k9! :smile:

    The inner product for vectors with complex numbers is defined with a complex conjugate (as tt showed).

    Without it, you have shown yourself that the resulting vector product does not satisfy the axioms for an inner product.
  5. Apr 28, 2012 #4

    I like Serena

    User Avatar
    Homework Helper

    It seems to me that you did not conjugate the first term.
    Perhaps you can write out your calculation?
  6. Apr 28, 2012 #5


    User Avatar

    Thanks for the welcomes, I see now I didn't read the definiton correctly and I was only trying to conjugate the first element in the vector, rather than the first term in each of the products.
  7. Apr 28, 2012 #6

    Stephen Tashi

    User Avatar
    Science Advisor

    What do you mean by "the first term"? State the definition you are using.
  8. Apr 30, 2012 #7
    <v,v>=0 only when v=0

    the definition for inner-product is Ʃ(v[itex]_{j}[/itex])([itex]\overline{v_{j}}[/itex]) for 1≤j≤n where n is the length of vector v
    note that [itex]\overline{v_{j}}[/itex] is defined as the adjoint, or conjugate transpose

    when dealing in ℝ, you'll never get <v,v>=0 because it is merely taking the square of each term {(a[itex]_{1}[/itex])[itex]^{2}[/itex]+...+(a[itex]_{k}[/itex])[itex]^{2}[/itex]}
    for all k[itex]\epsilon[/itex]dim(v) and then taking the square root of that sum
    √(Ʃa[itex]_k{}[/itex]) which means that the inside sum must be ≥0 else the inner product wouldn't exist because we are in ℝ,
    but would also never equal zero unless v=0 and then 0[itex]^{2}[/itex]=0

    it's the same for ℂ since squaring terms ends up those new terms becoming positive, rather non-zero and non-negative
  9. May 1, 2012 #8


    User Avatar
    Science Advisor
    Homework Helper

    Welcome to PF!

    Hi Krovski! Welcome to PF! :smile:
    Ah, no … that fooled me at first, too …

    it's "<v,v> >= 0" …

    and the >= is because 9k9 doesn't have a Mac with a key that types "≥" !! :biggrin:
  10. May 1, 2012 #9
    Re: Welcome to PF!

    good catch and thank you
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook