Complexification of a vector space

Sea V un espacio vectorial real, y J:V→V un R-isomorfismo tal que J2=−Id. Entonces J es llamada una complex structure en V. Cuando en un espacio vectorial real tenemos una tal estructura definida, podemos ver ese conjunto V como un espacio vectorial complejo de la siguiente manera:

(α+iβ)⋅v=αv+βJ(v)

Es fácil comprobar que con esta operacion V es un C-espacio vectorial. We will denote it by VJ.

Recíprocamente, si V es un espacio vectorial complejo, se puede considerar como un espacio vectorial real restringiendo los escalares. Y ademas, el automorfismo consistente en multiplicar por i es una estructura compleja.

En particular, consideremos Cn. Si lo identificamos con el R-espacio vectorial R2n, la multiplicación por i induce la estructura compleja

J(x1,y1,x2,y2,⋯)=(−y1,x1,−y2,x2,⋯)

cuya matriz en la base canónica es

M=(0−100⋯1000⋯000−1⋯0010⋯⋯⋯⋯⋯⋯)

con bloques

(0−110)

en la diagonal.

El automorfismo J recibe el nombre de estructura compleja estándar, y es única en el siguiente sentido:

Lema

Dada cualquier estructura compleja J1 de R2n, existe una matriz regular A tal que M1=A⋅M⋅A−1, donde estamos denotando por M y M1 a las matrices de los morfismos J y J1 en la base canónica de R2n.

Proof

Llamemos C1n al espacio vectorial complejo originado por R2n junto con la estructura compleja J1, y sea {f1,f2,⋯} una base de dicho C-espacio vectorial. Entonces

{f1,J1(f1),f2,J1(f2),⋯}

es también base de R2n, con la particularidad de que en dicha base la matriz de J1 es

(0−100⋯1000⋯000−1⋯0010⋯⋯⋯⋯⋯⋯)

Si A es la matriz del cambio de base se tiene que

M1=A⋅M⋅A−1.

◼

Así pues, el espacio cociente GL(2n,R)GL(n,C) determina todas las estructuras complejas de R2n mediante la aplicación [A]→A⋅M⋅A−1.
Para comprobarlo primero necesitaremos el siguiente

Lema

Sea V un espacio vectorial real con una estructura compleja J, y h una aplicación R-lineal de V en si mismo. El endomorfismo h es C-lineal si y sólo si h(Jv)=Jh(v).

Proof

Trivial. ◼

Así pues, consideremos en GL(2n,R)GL(n,C) las clases [A]=[B], entonces las matrices AMA−1 y BMB−1 son iguales, pues AB−1∈GL(n,C) y por tanto conmuta con M, luego

AMA−1=AM(A−1B)(B−1A)A−1==A(A−1B)M(B−1A)A−1=BMB−1.

Definition

Given a real vector space V, we call the complexification of V, and denote it by VC, to the set V⊕V with the complex structure given by

J : V \oplus V \rightarrow V \oplus V $$such that $J(v, w)=(-w, v)$. Obviously, $V^{\mathbb{C}}$ can be treated like a complex vector space. This can be constructed, alternatively, using tensor product. See the document in Calibre Complexification, complex structures, inner products, symplectic forms and linear differential equations'' for a good outline of this subject. ------ complex eigenvalues --- # Complexification and conjugation Given a vector space $V$ and its complexificated $V^{\mathbb{C}}$, we have a map $C$ (conjugation) that is involutive, $C^2=Id$, and such that the subspace of $V^{\mathbb{C}}$ sucht that

C(v)=v

isjusttheproper$V$.But,givenanyothermaplike$C$,say$D$,thesubspacesuchthat

D(v)=v

is a real vector space, call it $W$, such that $V^{\mathbb{C}}=W^{\mathbb{C}}$. See \begin{verbatim} https://math.stackexchange.com/questions/1061823/conjugation-map-on-a-complex-vector-space \end{verbatim} and TRTR page 414. ![Pasted image 20211230100427.png](/img/user/imagenes/Pasted%20image%2020211230100427.png)