ncg

bachelorthesis in physics
git clone git://popovic.xyz/ncg.git
Log | Files | Refs

commit 6410158b96d6b52ecbdc7285c54f986fd25af4fd
parent 08728ff7db193b88e1a312f16cd0404357269f9e
Author: miksa234 <milutin@popovic.xyz>
Date:   Wed, 17 Mar 2021 12:30:11 +0100

finishing up the worked exercises

Diffstat:
Mweek5.tex | 111+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++--
1 file changed, 109 insertions(+), 2 deletions(-)

diff --git a/week5.tex b/week5.tex @@ -227,6 +227,51 @@ exercises. differential operators` given $A$, that act on $H$? \end{question} Then there is a map $d:A\rightarrow \Omega _D ^1 (A)$, $d = [D, \cdot]$. +\begin{exercise} + Verify that 'd` is a derivation of the C* algebra + \begin{align*} + d(ab) = d(a)b + ad(b) \\ + d(a^*) = -d(a)^* + \end{align*} +\end{exercise} +\begin{solution} + For the record $d(\cdot) = [D, \cdot]$, then we have + \begin{enumerate} + \item + \begin{align*} + d(ab) &= [D, ab] = [D, a]b + a[D,b]\\ + &= d(a)b + ad(b) + \end{align*} + \item + \begin{align*} + d(a^*) &= [D, a^*] = Da^* - a^*D \\ + &=-(D^*a - aD^*) = -[D^*, a] \\ + &= -d(a)^* + \end{align*} + \end{enumerate} +\end{solution} +\begin{exercise} + Verify that $\Omega _D^1 (A)$ is an $A$-bimodule by rewriting + \begin{align*} + a(a_k[D, b_k]b = \sum_k a'_k[D, b'_k] \;\;\;\; a'_k, b'_k \in A + \end{align*} +\end{exercise} +\begin{solution} + First off we know the algebra is associative then we know that elements + in $A$ can be represented faithfully on a Hilbert space $H$. Because of + the Hilbert Basis $\{\textbf{n}_i\}_{i\in \mathbb{N}}$ of the Hilbert space we can decompose these elements + in therms of the basis elements. + \begin{align*} + aa_k &= \sum _{\textbf{n}}(\langle a, \textbf{n} \rangle) a_k \\ + &= \sum _{k} a'_{k} + \end{align*} + Which would than be the same as the sum of some elements + $\a'_{k} \in A$. Then we calculate the commutator: + \begin{align*} + [D, b_k] b = d(b_k)b = d(b_kb) - b_kd(b)\\ + \end{align*} +I don't think this is correct I'll try it again +\end{solution} \begin{lemma} Let $(A, H, D) = (M_n(\mathbb{C}, \mathbb{C}^n, D)$, with $D$ a hermitian @@ -240,14 +285,28 @@ Then there is a map $d:A\rightarrow \Omega _D ^1 (A)$, $d = [D, \cdot]$. Assume $D = \sum _i \lambda _i e_{ii}$ (diagonal), $\lambda _i \in \mathbb{R}$ and $\{e_{ij}\}$ the basis of $M_n(\mathbb{C}$. For fixed $i$, $j$ choose $k$ such that $\lambda _k \neq \lambda _j$ then - \begin{align*} + \begin{align} \label{basis} \left(\frac{1}{\lambda _k - \lambda _j} e_{ik}\right) [D, e_{kj}] = e_{ij} - \end{align*} + \end{align} $e_{ij}\in \Omega _D ^1 (A)$ by the above definition. And $\Omega _D ^1 (A) \subset L(\mathbb{C}^n) = H \simeq M_n(\mathbb{C}) = A$ \end{proof} +\begin{exercise} + Consider $(A=\mathbb{C}^2, H=\mathbb{C}^2, + D = \begin{pmatrix} 0 & \lambda \\ \bar{\lambda} & 0 + \end{pmatrix})$ with $\lambda \neq 0$. Show that $\Omega _D^1(A) + \simeq M_2(\mathbb{C})$ +\end{exercise} +\begin{solution} + Because of the Hilbert Basis $D$ can be extended in terms of + the basis of $M_2(\mathbb{C})$, plugging this into Equation + \ref{basis} will get us the same cyclic result, thus + $\Omega _D^1(A) \simeq M_2(\mathbb{C})$ +\ +\end{solution} + \subsection{Morphisms Between Finite Spectral Triples} \begin{definition} two finite spectral tripes $(A_1, H_1, D_1)$ and $(A_2, H_2, D_2)$ are @@ -272,6 +331,54 @@ Some remarks $\Omega _D^1 (A)$. \end{itemize} +\begin{exercise} + Show that the unitary equivalence between finite spectral + triples is a equivalence relation +\end{exercise} + +\begin{solution} + An equivalence relation needs to satisfy reflexivity, symmetry + transitivity. + Let $(A_1, H_1, D_1)$, $(A_2, H_2, D_2)$ and $(A_3, H_3, D_3)$ + be three finite spectral triples. + \newline + + For reflexivity $(A_1, H_1, D_1) \sim (A_1, H_1, D_1)$. So there + exists a $U: H_1 \rightarrow H_1$ unitary, which is the identity + and always exists. + \newline + + For symmetry we need + \begin{align*} + (A_1, H_1, D_1) \sim (A_2, H_2, D_2) \Leftrightarrow + (A_2, H_2, D_2) \sim (A_1, H_1, D_1) + \end{align*} + because $U$ is unitary: + \begin{align*} + &U\pi_1(a)U^* = \pi_2(a) \;\;\; | \cdot U^*\boxdot U \\ + &U^*U\pi_1(a)U^*U = \pi_1(a) = U^*\pi_2(a)U \\ + \end{align*} + The same with the symmetric operator $D$. + + \newline + For transitivity we need + \begin{align*} + (A_1, H_1, D_1) &\sim (A_2, H_2, D_2) \;\;\; \text{and} \;\;\; + (A_2, H_2, D_2) \sim (A_3, H_3, D_3) \\ + &\Rightarrow (A_1, H_1, D_1) \sim (A_3, H_3, D_3) + \end{align*} + There are two unitary maps $U_{12}:H_1 \rightarrow H_2$ and + $U_{23}: H_2 \rightarrow H_3$ then + \begin{align*} + U_{23}U_{12} \pi_1(a) U^*_{12}U^*_{23} &= U_{23} + \pi_2(a) U_23^* \\ + &= \pi_3(a) \\ + U_{23}U_{12} D_1U^*_{12}U^*_{23} &= U_{23} + D_2 U_23^* \\ + &= D_3 + \end{align*} +\end{solution} + Extending the this relation we look again at the notion of equivalence from Morita equivalence of Matrix Algebras. \newline