2
$\begingroup$

My textbook says that to find the ket $|ψ\rangle$ in the same position basis as the ket $|ø\rangle$ we do the following: $$|ψ\rangle=\int dø|ø\rangle \langle ø|ψ\rangle$$ Firstly can $|ø\rangle$ be any ket? i.e. this expression just puts $|ψ\rangle$ in the same basis as $|ø\rangle$ regardless of the components of $|ø\rangle$?

Secondly my textbook goes on to say to place $|ψ\rangle$ in the position basis we do the following: $$|ψ\rangle=\int d^3r\ |\mathbf{r}\rangle\langle \mathbf{r}|ψ\rangle$$ Why have we suddenly gained a cubed sign?

Are we taking the integral over nothing? i.e. are the integrals we are doing simply $\int dø$ and $\int d^3r$?

(I am new to this sort of physics/maths and am self teaching so please can you keep the explanations relativity simple) thanks

$\endgroup$
1
  • $\begingroup$ That is because the Hilbert space you are considering is $L^2(\mathbb{R}^3)$, the space of square-integral functions of $\mathbb{R}^3$. I've replaced $r$ by $\mathbf{r}$ to emphasise this in your question. $\endgroup$
    – suresh
    Commented Jul 24, 2014 at 7:38

2 Answers 2

1
$\begingroup$

So first of all, the first equation you gave is only correct, if the $|ø\rangle$ form a basis. It has nothing to do with "in which basis they are".

The easiest way to understand this is probably with a 3D vector-analogy. So if $b_i$, $i=1\dots3$ form a basis, for any vector $v$ it is legitimate to write $$v=\sum_{i=1}^3 b_i (b_i\cdot v)$$ There, the $b_i\cdot v$ are the components of $v$ in the representation of the $b_i$.

It is the very same for bras and kets. It is "just" not 3d but has infinite dimension, so if we have a basis of infinite $|ø\rangle$, $ø\in \mathbb{R}$ or $|r\rangle$, $r\in \mathbb{R}^3$, denote the scalar product ($a\cdot b$) using dirac notatiton ($\langle a|b\rangle$), and write integrals instead of sums we get te formulas given by you (mathematically this is non trivial). Therefor the $\langle r|\psi\rangle$ are the components in the position basis.

$\endgroup$
6
  • $\begingroup$ Would $<r|=(i,j,k)$ ? $\endgroup$
    – user43487
    Commented Jul 24, 2014 at 8:50
  • $\begingroup$ $\langle r|$ is a function (it is element of the dual space) acting on a vector/ket giving a scalar. So yes, in this simplified picture you are correct. $\endgroup$
    – Hagadol
    Commented Jul 24, 2014 at 8:56
  • $\begingroup$ Is it the fact that $<r|$ is 3 dimensional that we have the cubed? thanks $\endgroup$
    – user43487
    Commented Jul 24, 2014 at 9:00
  • $\begingroup$ If you have three dimensional space, you need all those three dimensions to form a complete basis. So yes, that is the reason. $\endgroup$
    – M.Herzkamp
    Commented Jul 24, 2014 at 10:19
  • $\begingroup$ I know this is a simple question but lets say $<A|$ is a basis consisting of $(a_1,a_1,...,a_n)$ then any vector $|V>$ can be written as $|V>=v_ia_i=<A|V>$ where $a_i$ themsevles are vectors?? $\endgroup$
    – user43487
    Commented Jul 24, 2014 at 11:25
1
$\begingroup$

Regarding the first part of your question,they have just inserted a complete set of basis because $|\phi>$ is a basis in some infinite dimensional Hilbert space (in your case), therefore sum (integral) of all such bases is identity on the Hilbert space. Note that in second part

$\langle r|\phi\rangle$=$\phi(r)$.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.