Why is <v,v> >= 0 ?


by 9k9
Tags: <v, v>
9k9
9k9 is offline
#1
Apr27-12, 04:28 PM
P: 5
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:

v=(1,2i)

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?
Phys.Org News Partner Science news on Phys.org
Better thermal-imaging lens from waste sulfur
Hackathon team's GoogolPlex gives Siri extra powers
Bright points in Sun's atmosphere mark patterns deep in its interior
tiny-tim
tiny-tim is offline
#2
Apr27-12, 05:59 PM
Sci Advisor
HW Helper
Thanks
tiny-tim's Avatar
P: 26,167
hi 9k9! welcome to pf!

(1,2i).(1,-2i) = 1.1 + 2i.-2i = 1 - 4i2 = 5
I like Serena
I like Serena is offline
#3
Apr28-12, 05:31 AM
HW Helper
I like Serena's Avatar
P: 6,189
Welcome to PF, 9k9!

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.

I like Serena
I like Serena is offline
#4
Apr28-12, 10:15 AM
HW Helper
I like Serena's Avatar
P: 6,189

Why is <v,v> >= 0 ?


It seems to me that you did not conjugate the first term.
Perhaps you can write out your calculation?
9k9
9k9 is offline
#5
Apr28-12, 10:17 AM
P: 5
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.
Stephen Tashi
Stephen Tashi is offline
#6
Apr28-12, 11:10 AM
Sci Advisor
P: 3,172
Quote Quote by 9k9 View Post
Using the definition where you do the dot product, while conjugating the first term
What do you mean by "the first term"? State the definition you are using.
Krovski
Krovski is offline
#7
Apr30-12, 09:29 PM
P: 11
<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
tiny-tim
tiny-tim is offline
#8
May1-12, 03:37 AM
Sci Advisor
HW Helper
Thanks
tiny-tim's Avatar
P: 26,167
Hi Krovski! Welcome to PF!
Quote Quote by Krovski View Post
<v,v>=0 only when v=0
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 "≥" !!
Krovski
Krovski is offline
#9
May1-12, 05:49 AM
P: 11
Quote Quote by tiny-tim View Post
Hi Krovski! Welcome to PF!


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 "≥" !!
good catch and thank you


Register to reply