I have doubts regarding a statement related to the adjoint linear transformation, here is the statement and its proof:

Proposition: Let $(V,\langle,\rangle)$ be a finite dimensional vector space with inner product and let $f:V \to V$ be a linear transformation. Let $B$ be an orthonormal basis of $V$. Then $|f^*|_B=(|f|_B)^* $

Proof: Suppose $B=\{v_1,…,v_n\}$ is an orthonormal basis given in $V$. Then, for each $1\leq i,j \leq n$, $$(|f^*|_B)_{ij}=\langle f^*(v_j),v_i \rangle=\overline {\langle v_i, f^*(v_j) \rangle}= (\overline {|f|_B})_{ji}=(\overline {{|f|_B}^*})_{ij}.$$

I am trying to understand these two equalities:

  1. $$(|f^*|_B)_{ij}=\langle f^*(v_j),v_i \rangle$$ (first equality)
  2. $$(\overline {|f|_B})_{ji}=(\overline {{|f|_B}^*})_{ij}.$$ (last equality)

With respect to 1), why the entry $ij$ is equal to $\langle f^*(v_j),v_i\rangle$, in other words, why the $i$-th coordinate of $f^*(v_j)$ with respect to the basis $B$ is $\langle f^*(v_j),v_i \rangle$? I can’t figure that out.

According to my linear algebra textbook, $f^*$ evaluated in $w \in V$ is $$f^*(w)=\sum_{i=1}^n \langle w,f(v_i) \rangle v_i$$ where $\{v_1,…,v_n\}$ is an orthonormal basis, I’ve tried to use this definition of self adjoint in order to explain myself equalities 1) and 2) but I couldn’t. I would appreciate if someone could explain these two to me.


why the $i$-th coordinate of $f^*(v_j)$ with respect to the basis $B$ is $\langle f^*(v_j),v_i \rangle$?

Because $B$ is an orthonormal basis. When a vector $v$ is expanded in the orthonormal basis $(e_k)$, i.e., $v=\sum c_k e_k$, taking the inner product of both sides with $e_i$ results in $\langle v,e_i\rangle = \sum c_k \delta_{ik} = c_i$.

Once you see this relation of matrix entries with inner product, the rest follows: $$\langle f^*(v_j),v_i \rangle = \langle v_j,f(v_i) \rangle \tag{*}$$ by the definition of an adjoint. Then the right hand side of (*) is related to the matrix representing $f$, as it was previously done for $f^*$.


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.