Trace in Terms of Orthonormal Basis

From ProofWiki
Jump to navigation Jump to search

Theorem

Let $\mathbb K \subset \C$ be a field.



Let $\struct {V, \innerprod {\, \cdot \,} {\, \cdot \,} }$ be an inner product space over $\mathbb K$ of dimension $n$.

Let $\tuple {e_1, \ldots, e_n}$ be an orthonormal basis of $V$.

Let $f: V \to V$ be a linear operator.


Then its trace equals:

$\map \tr f = \ds \sum_{i \mathop = 1}^n \innerprod {\map f {e_i} } {e_i}$


Proof

Let $\ds \map f {e_i} = \sum_{j \mathop = 1}^n c_{ij} e_j$

Let $A$ be the matrix relative to the basis $\tuple {e_1, \ldots, e_n}$.

Then by the above assumption, $A_{ij} = c_{ij}$.


Then:

\(\ds \map \tr f\) \(=\) \(\ds \map \tr A\) Definition of Trace of Linear Operator
\(\ds \) \(=\) \(\ds \sum_{i \mathop = 1}^n A_{ii}\) Definition of Trace of Matrix
\(\ds \) \(=\) \(\ds \sum_{i \mathop = 1}^n c_{ii}\) From above


Now it remains to show that $c_{ii} = \innerprod {\map f {e_i} } {e_i}$:

\(\ds \innerprod {\map f {e_i} } {e_i}\) \(=\) \(\ds \innerprod {\sum_{j \mathop = 1}^n c_{ij} e_j} {e_i}\) From above assumption
\(\ds \) \(=\) \(\ds \sum_{j \mathop = 1}^n c_{ij} \innerprod {e_j} {e_i}\) Axiom $(2)$ of Inner Product: Bilinearity
\(\ds \) \(=\) \(\ds c_{ii} (1)\) Definition of Orthonormal Subset: other terms vanish
\(\ds \) \(=\) \(\ds c_{ii}\)

$\blacksquare$


Also see