Matrice de comutare

În matematică, în special în algebra liniară, o matrice de comutare este folosită pentru transformarea formei vectorizate a unei matrice în forma vectorizată a transpusei. Mai exact, matricea de comutare K(m,n) este matricea nm × mn care, pentru orice matrice A m × n, transformă vec(A) în vec(AT):

K(m,n) vec(A) = vec(AT) .

Aici vec(A) este vectorul coloană mn × 1 obținut prin aranjarea coloanelor lui A una peste alta:

vec ( A ) = [ A 1 , 1 , , A m , 1 , A 1 , 2 , , A m , 2 , , A 1 , n , , A m , n ] T {\displaystyle \operatorname {vec} (\mathbf {A} )=[\mathbf {A} _{1,1},\ldots ,\mathbf {A} _{m,1},\mathbf {A} _{1,2},\ldots ,\mathbf {A} _{m,2},\ldots ,\mathbf {A} _{1,n},\ldots ,\mathbf {A} _{m,n}]^{\mathrm {T} }}

unde A = [Ai,j]. Cu alte cuvinte, vec(A) este vectorul obținut prin vectorizarea lui A în ordinea principală a coloanelor. Similar, vec(AT) este vectorul care se obține prin vectorizarea lui A în ordinea principală a liniilor.

Proprietăți

  • Matricea de comutare este un tip special de matrice de permutare, prin urmare, este ortogonală⁠(d). În special, K(m,n) este egală cu P π {\displaystyle \mathbf {P} _{\pi }} , unde π {\displaystyle \pi } este permutarea peste { 1 , , m n } {\displaystyle \{1,\dots ,mn\}} pentru care
π ( i + m ( j 1 ) ) = j + n ( i 1 ) , i = 1 , , m , j = 1 , , n . {\displaystyle \pi (i+m(j-1))=j+n(i-1),\quad i=1,\dots ,m,\quad j=1,\dots ,n.}
  • Înlocuirea lui A cu AT în definiția matricei de comutare arată că K (m,n) = (K(n,m))T. Prin urmare, în cazul particular al lui m = n matricea de comutare este o involuție și o matrice simetrică.
  • Utilizarea principală a matricei de comutare și sursa numelui acesteia este de a comuta produsul Kronecker: pentru orice matrice A m × n și orice matrice B r × q,
K ( r , m ) ( A B ) K ( n , q ) = B A . {\displaystyle \mathbf {K} ^{(r,m)}(\mathbf {A} \otimes \mathbf {B} )\mathbf {K} ^{(n,q)}=\mathbf {B} \otimes \mathbf {A} .}
Această proprietate este adesea folosită în dezvoltarea statisticilor de ordin superior ale matricelor de covarianță Wishart.[1]
  • Cazul lui n = q = 1 pentru ecuația de mai sus arată că pentru orice vectori coloană v,w de dimensiuni m,r există
K ( r , m ) ( v w ) = w v . {\displaystyle \mathbf {K} ^{(r,m)}(\mathbf {v} \otimes \mathbf {w} )=\mathbf {w} \otimes \mathbf {v} .}
Această proprietate este motivul pentru care această matrice este denumită operator de schimb în contextul teoriei informației cuantice.
  • Două forme explicite pentru matricea de comutare sunt următoarele: dacă prin er,j se notează al j-lea vector canonic al dimensiunii „r” (adică vectorul cu 1 în a j-a coordonată și 0 în rest) atunci
K ( r , m ) = i = 1 r j = 1 m ( e r , i e m , j T ) ( e m , j e r , i T ) = i = 1 r j = 1 m ( e r , i e m , j ) ( e m , j e r , i ) T . {\displaystyle \mathbf {K} ^{(r,m)}=\sum _{i=1}^{r}\sum _{j=1}^{m}\left(\mathbf {e} _{r,i}{\mathbf {e} _{m,j}}^{\mathrm {T} }\right)\otimes \left(\mathbf {e} _{m,j}{\mathbf {e} _{r,i}}^{\mathrm {T} }\right)=\sum _{i=1}^{r}\sum _{j=1}^{m}\left(\mathbf {e} _{r,i}\otimes \mathbf {e} _{m,j}\right)\left(\mathbf {e} _{m,j}\otimes \mathbf {e} _{r,i}\right)^{\mathrm {T} }.}
  • Matricea de comutare poate fi exprimată ca următoarea matrice de blocuri:
K ( m , n ) = [ K 1 , 1 K 1 , n K m , 1 K m , n , ] , {\displaystyle \mathbf {K} ^{(m,n)}={\begin{bmatrix}\mathbf {K} _{1,1}&\cdots &\mathbf {K} _{1,n}\\\vdots &\ddots &\vdots \\\mathbf {K} _{m,1}&\cdots &\mathbf {K} _{m,n},\end{bmatrix}},}
unde elementul p,q al matricei de blocuri Ki,j n × m este dat de
K i j ( p , q ) = { 1 i = q  si  j = p , 0 altfel . {\displaystyle \mathbf {K} _{ij}(p,q)={\begin{cases}1&i=q{\text{ si }}j=p,\\0&{\text{altfel}}.\end{cases}}}
De exemplu,
K ( 3 , 4 ) = [ 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 ] . {\displaystyle \mathbf {K} ^{(3,4)}=\left[{\begin{array}{ccc|ccc|ccc|ccc}1&0&0&0&0&0&0&0&0&0&0&0\\0&0&0&1&0&0&0&0&0&0&0&0\\0&0&0&0&0&0&1&0&0&0&0&0\\0&0&0&0&0&0&0&0&0&1&0&0\\\hline 0&1&0&0&0&0&0&0&0&0&0&0\\0&0&0&0&1&0&0&0&0&0&0&0\\0&0&0&0&0&0&0&1&0&0&0&0\\0&0&0&0&0&0&0&0&0&0&1&0\\\hline 0&0&1&0&0&0&0&0&0&0&0&0\\0&0&0&0&0&1&0&0&0&0&0&0\\0&0&0&0&0&0&0&0&1&0&0&0\\0&0&0&0&0&0&0&0&0&0&0&1\end{array}}\right].}

Exemplu

Fie A {\displaystyle A} matricea 3 × 2 {\displaystyle 3\times 2} :

A = [ 1 4 2 5 3 6 ] . {\displaystyle A={\begin{bmatrix}1&4\\2&5\\3&6\\\end{bmatrix}}.}

A {\displaystyle A} are vectorizarea în ordinea princpală a coloanelor, respectiv în ordinea princpală a liniilor:

v col = vec ( A ) = [ 1 2 3 4 5 6 ] , v row = vec ( A T ) = [ 1 4 2 5 3 6 ] . {\displaystyle \mathbf {v} _{\text{col}}=\operatorname {vec} (A)={\begin{bmatrix}1\\2\\3\\4\\5\\6\\\end{bmatrix}},\quad \mathbf {v} _{\text{row}}=\operatorname {vec} (A^{\mathrm {T} })={\begin{bmatrix}1\\4\\2\\5\\3\\6\\\end{bmatrix}}.}

Matricea de comutare asociată este

K = K ( 3 , 2 ) = [ 1 1 1 1 1 1 ] , {\displaystyle K=\mathbf {K} ^{(3,2)}={\begin{bmatrix}1&\cdot &\cdot &\cdot &\cdot &\cdot \\\cdot &\cdot &\cdot &1&\cdot &\cdot \\\cdot &1&\cdot &\cdot &\cdot &\cdot \\\cdot &\cdot &\cdot &\cdot &1&\cdot \\\cdot &\cdot &1&\cdot &\cdot &\cdot \\\cdot &\cdot &\cdot &\cdot &\cdot &1\\\end{bmatrix}},}

(unde orice {\displaystyle \cdot } indică un zero). După cum este de așteptat, sunt valabile relațiile:

K T K = K K T = I 6 {\displaystyle K^{\mathrm {T} }K=KK^{\mathrm {T} }=\mathbf {I} _{6}}
K v col = v row {\displaystyle K\mathbf {v} _{\text{col}}=\mathbf {v} _{\text{row}}}

Note

  1. ^ en von Rosen, Dietrich (). „Moments for the Inverted Wishart Distribution”. Scand. J. Stat. 15: 97–109. 

Bibliografie

  • Jan R. Magnus and Heinz Neudecker (1988), Matrix Differential Calculus with Applications in Statistics and Econometrics, Wiley.
Portal icon Portal Matematică