In my textbook there is the following example of how to find an orthonomal basis with Gram-Schmidt process.

Let P2, the vector space of all polynomials of degree two or less, with the inner product defined as: $$<p, q> = p(−1)q(−1) + p(0)q(0) + p(1)q(1)\, for\ p, q\in P_2 )$$

Find an orthonormal basis for P2.

The basis is given by {$1,X,X^2$}. Let $V$ be an inner product space with the basis $V = (v_1,v_2,...v_n)$ and let $p_k$ for $k = 1,2,3,...n-1$ is the orthogonal projection of $v_{k+1}$ on the subspace $Span(v_1,v_2,...v_k)$. Then we have that $W = (v_1,v_2-p_1,v_3-p_2,...,v_n-p_{n-1})$ is the orthogonal basis for V.

$u_1 = \frac {1}{{\left\|v_1\right\|}}v_1 = \frac {1}{\sqrt3}\cdot 1$

It is now possible to find $p_1 = <x,\frac {1}{\sqrt3}>\cdot \frac{1}{\sqrt3}=0$...

And then it continues to calculate the remaning $u_k$'s and $p_k$'s. However, my question is just about the computation itself.

So for $u_1$ the norm of $v_1$ is calculated by $\sqrt{<1,1>} = \sqrt{1\cdot1 + 1\cdot1 + 1\cdot1} = \sqrt3$ and if I had to the it for $X$ then it would be $\sqrt{<X,X>} = \sqrt{1+0+1} = \sqrt2$ right? What I do not understand is, how $v_1$ is simply 1 and for $p_1$ how does $<x,\frac{1}{\sqrt3}> \cdot \frac{1}{\sqrt3}$ = 0 ?

Hope someone can clarify


I don't understand your first question. I'd say that $v_1$ is simply $1$ because whoever created that exercise decide to start with the basis $(v_1,v_2,v_3)=(1,X,X^2)$.

And $\left\langle X,\frac1{\sqrt3}\right\rangle=(-1)\times\frac1{\sqrt3}+0\times\frac1{\sqrt3}+1\times\frac1{\sqrt3}=0$.

  • $\begingroup$ That makes sense, thank you! For the first question I was just confused why it was $1$ as it is done with the basis you described. $\endgroup$ May 19 '18 at 14:24

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.