Math 540 - Solutions to Selcted HW Problems - Due Feb 6



Problem # A.2: An \(m\)-by-\(n\) matrix with coefficients in a field \(\mathbb{F}\) is defined to be an \(m\)-by-\(n\) array of elements of \(\mathbb{F}\). We write \(M_{m,n}(\mathbb{F})\) for the set of all such matrices, so an element \(A\in M_{m,n}(\mathbb{F})\) looks like \[ A = \begin{pmatrix} a_{11} & a_{12} & \cdots & a_{1n} \\ a_{21} & a_{22} & \cdots & a_{2n} \\ \vdots & & \ddots & \vdots \\ a_{m1} & a_{m2} & \cdots & a_{mn} \\ \end{pmatrix} \] We make \(M_{m,n}(\mathbb{F})\) into a vector space in the obvious way: \[ \begin{pmatrix} a_{11} & \cdots & a_{1n} \\ \vdots & \ddots & \vdots \\ a_{m1} & \cdots & a_{mn} \\ \end{pmatrix} + \begin{pmatrix} b_{11} & \cdots & b_{1n} \\ \vdots & \ddots & \vdots \\ b_{m1} & \cdots & b_{mn} \\ \end{pmatrix} = \begin{pmatrix} a_{11}+b_{11} & \cdots & a_{1n}+b_{1n} \\ \vdots & \ddots & \vdots \\ a_{m1}+b_{m1} & \cdots & a_{mn}+b_{mn} \\ \end{pmatrix} \] and \[ c \begin{pmatrix} a_{11} & \cdots & a_{1n} \\ \vdots & \ddots & \vdots \\ a_{m1} & \cdots & a_{mn} \\ \end{pmatrix} = \begin{pmatrix} ca_{11} & \cdots & ca_{1n} \\ \vdots & \ddots & \vdots \\ ca_{m1} & \cdots & ca_{mn} \\ \end{pmatrix} \]
(a) Write down a basis for the \(\mathbb{F}\)-vector space \(M_{3,2}(\mathbb{F})\) of 3-by-2 matrices. What is the dimension of \(M_{3,2}(\mathbb{F})\)?
(b) More generally, what is the dimension of \(M_{m,n}(\mathbb{F})\)?
Solution: (a) The following six vectors form a basis for \(M_{3,2}(\mathbb{F})\), so \(M_{3,2}(\mathbb{F})\) has dimension 6. \[ \begin{pmatrix} 1 & 0 \\ 0 & 0 \\ 0 & 0 \\ \end{pmatrix},\quad \begin{pmatrix} 0 & 1 \\ 0 & 0 \\ 0 & 0 \\ \end{pmatrix},\quad \begin{pmatrix} 0 & 0 \\ 1 & 0 \\ 0 & 0 \\ \end{pmatrix},\quad \begin{pmatrix} 0 & 0 \\ 0 & 1 \\ 0 & 0 \\ \end{pmatrix},\quad \begin{pmatrix} 0 & 0 \\ 0 & 0 \\ 1 & 0 \\ \end{pmatrix},\quad \begin{pmatrix} 0 & 0 \\ 0 & 0 \\ 0 & 1 \\ \end{pmatrix}. \] To see this, note that \[ \begin{pmatrix} a & b \\ c & d \\ e & f \\ \end{pmatrix}= a\begin{pmatrix} 1 & 0 \\ 0 & 0 \\ 0 & 0 \\ \end{pmatrix}+ b\begin{pmatrix} 0 & 1 \\ 0 & 0 \\ 0 & 0 \\ \end{pmatrix}+ c\begin{pmatrix} 0 & 0 \\ 1 & 0 \\ 0 & 0 \\ \end{pmatrix}+ d\begin{pmatrix} 0 & 0 \\ 0 & 1 \\ 0 & 0 \\ \end{pmatrix}+ e\begin{pmatrix} 0 & 0 \\ 0 & 0 \\ 1 & 0 \\ \end{pmatrix}+ f\begin{pmatrix} 0 & 0 \\ 0 & 0 \\ 0 & 1 \\ \end{pmatrix}. \] This shows that every matrix in \(M_{3,2}(\mathbb{F})\) is a linear combination of the given 6 matrices, so the given 6 matrices span. Further, it shows that the only linear combination of the 6 matrices that equals the 0 matrix is by taking the 0 multiple of each of them, so the 6 matrices are linearly indepedent.
(b) Let \(E_{ij}\) be the matrix that has a \(1\) for its entry in the \(i\)'th row and \(j\)'th column, and every other entry is 0. Then \[ \{ E_{ij} : 1\le i\le m,\; 1\le j\le n \} \] is a basis for \(M_{m,n}(\mathbb{F})\), so \(M_{m,n}(\mathbb{F})\) has dimension \(mn\). To see that this is a basis, we note that \[ \sum_{i=1}^m \sum_{j=1}^n a_{ij}E_{ij} = \begin{pmatrix} a_{11} & a_{12} & \cdots & a_{1n} \\ a_{21} & a_{22} & \cdots & a_{2n} \\ \vdots & & \ddots & \vdots \\ a_{m1} & a_{m2} & \cdots & a_{mn} \\ \end{pmatrix}, \] so we get every matrix in \(M_{m,n}(\mathbb{F})\) as a linear combination of the \(E_{ij}\) matrices, so the \(E_{ij}\) matrices span, and the only linear combination of the \(E_{ij}\) matrices that equals the zero matrix is taking all of the \(a_{ij}\) to be 0, so the \(E_{ij}\) matrices are linearly independent.


Problem # A.3: The transpose of a (square) matrix \(A\), denoted \(A^*\), is obtained by flipping the entries across the main diagonal. So for example \[ \begin{pmatrix} 1&2&3\\ 4&5&6\\ 7&8&9\\ \end{pmatrix}^* = \begin{pmatrix} 1&4&7\\ 2&5&8\\ 3&6&9\\ \end{pmatrix}. \] A matrix \(A\) is symmetric if \(A^*=A\) and it is anti-symmetric if \(A^*=-A\).
(a) Prove that the set of \(n\)-by-\(n\) symmetric matrices is a vector subspace of \(M_{n,n}(\mathbb{F})\).
(b) Find a basis for the space of 2-by-2 symmetric matrices. What is its dimension?
(c) Generalize by describing a basis for the space of \(n\)-by-\(n\) symmetric matrices and computing its dimension. (It might help to start with 3-by-3.)
(d) Prove that the set of \(n\)-by-\(n\) anti-symmetric matrices is also a vector subspace of \(M_{n,n}(\mathbb{F})\), describe a basis, and compute its dimension.
(e) (Bonus) Let's write \(M_{n,n}(\mathbb{F})^{\text{sym}}\) for the space of symmetric matrices and \(M_{n,n}(\mathbb{F})^{\text{anti-sym}}\) for the space of anti-symmetric matrices. Prove that \[ M_{n,n}(\mathbb{F}) = M_{n,n}(\mathbb{F})^{\text{sym}} + M_{n,n}(\mathbb{F})^{\text{anti-sym}}. \] Is this a direct sum of vector spaces?
Solution: What happens if we add two matrices and then take their transpose? We compute \[ \begin{align*} (A+B)^* &= \left(\begin{pmatrix} a_{11} & \cdots & a_{1n} \\ \vdots & \ddots & \vdots \\ a_{m1} & \cdots & a_{mn} \\ \end{pmatrix} + \begin{pmatrix} b_{11} & \cdots & b_{1n} \\ \vdots & \ddots & \vdots \\ b_{m1} & \cdots & b_{mn} \\ \end{pmatrix}\right)^* \\ &= \begin{pmatrix} a_{11}+b_{11} & \cdots & a_{1n}+b_{1n} \\ \vdots & \ddots & \vdots \\ a_{m1}+b_{m1} & \cdots & a_{mn}+b_{mn} \\ \end{pmatrix}^* \\ &= \begin{pmatrix} a_{11}+b_{11} & \cdots & a_{m1}+b_{m1}\\ \vdots & \ddots & \vdots \\ a_{1n}+b_{1n} & \cdots & a_{mn}+b_{mn} \\ \end{pmatrix} \\ &= \left(\begin{pmatrix} a_{11} & \cdots & a_{m1} \\ \vdots & \ddots & \vdots \\ a_{1n} & \cdots & a_{mn} \\ \end{pmatrix} + \begin{pmatrix} b_{11} & \cdots & b_{m1} \\ \vdots & \ddots & \vdots \\ b_{1n} & \cdots & b_{mn} \\ \end{pmatrix}\right) \\ &= A^* + B^* \end{align*} \] and similarly \((cA)^* = c A^*\). So we have the useful formulas \[ (A+B)^*=A^*+B^*\quad\text{and}\quad(cA)^* = c A^*. \]
(a) Suppose that \(A\) and \(B\) are symmetric matrices, so \(A^*=A\) and \(B^*=B\). Then using the formulas that we just proved, we have \[ (A+B)^* = A^*+B^* = A+B \quad\text{and}\quad (cA)^* = cA^* = cA. \] This shows that the set of symmetric matrices \(M_{n,n}^{\text{sym}}(\mathbb{F})\) is closed under addition and scalar multiplication, so it is a vector subspace of \(M_{n,n}(\mathbb{F})\).
(b) Consider the three matrices \[ E_{11} = \begin{pmatrix} 1&0\\ 0&0\\ \end{pmatrix},\quad E_{22} = \begin{pmatrix} 0&0\\ 0&1\\ \end{pmatrix},\quad F_{12} = \begin{pmatrix} 0&1\\ 1&0\\ \end{pmatrix}. \] Then any 2-by-2 symmetric matrix can be written as \[ \begin{pmatrix} a&b\\ b&d\\ \end{pmatrix} = \begin{pmatrix} a&0\\ 0&0\\ \end{pmatrix} + \begin{pmatrix} 0&0\\ 0&d\\ \end{pmatrix} + \begin{pmatrix} 0&b\\ b&0\\ \end{pmatrix} = aE_{11} + dE_{22} + bF_{12} \] in exactly one way, so \(\{E_{11},E_{22},F_{12}\}\) is a basis for \(M_{n,n}^{\text{sym}}(\mathbb{F})\). In particular, \(\dim M_{n,n}^{\text{sym}}(\mathbb{F})=3\).
(c) More generally, as in Problem A.2, let \(E_{ij}\) be the matrix that has a \(1\) for its entry in the \(i\)'th row and \(j\)'th column, and every other entry is 0. Also let \[ F_{ij} = E_{ij} + E_{ji}, \] so \(F_{ij}\) has two 1's and the rest of its entries are 0. Since it's clear that \[ E_{ij}^* = E_{ji}, \quad\text{we have}\quad F_{ij}^* = (E_{ij} + E_{ji})^* = E_{ij}^* + E_{ji}^* = E_{ji} + E_{ij} = F_{ij}, \] so \(F_{ij}\) is a symmetric matrix. It's also clear that \(E_{ii}\) is a symmetric matrix. Then generalizing the proof in (b), one sees that \[ \{E_{ii} : 1\le i\le n\} \cup \{F_{ij} : 1\le i\lt j\le n\} \] is a basis for \(M_{n,n}^{\text{sym}}(\mathbb{F})\). (Note that \(F_{ij}=F_{ji}\), so we only need one of them.) More precisely, since a symmetric matrix has \(a_{ij}=a_{ji}\), we have \[ \begin{pmatrix} a_{11} & a_{12} & \cdots & a_{1n} \\ a_{21} & a_{22} & \cdots & a_{2n} \\ \vdots & & \ddots & \vdots \\ a_{n1} & a_{n2} & \cdots & a_{nn} \\ \end{pmatrix} = \sum_{i=1}^n a_{ii}E_{ii} + \sum_{i=1}^{n-1} \sum_{j=i+1}^n a_{ij}F_{ij}. \] Finally, \[ \dim M_{n,n}^{\text{sym}}(\mathbb{F}) = n + \frac{n(n-1)}{2} = \frac{n^2+n}{2}. \]
(d) This is similar, but instead of using the \(E_{ij}\) and the \(F_{ij}\), use the matrices \(G_{ij}=E_{ij}-E_{ji}\). Note that \[ G_{ij}^* = (E_{ij}-E_{ji})^* = E_{ij}^*-E_{ji}^* = E_{ji}-E_{ij} = -G_{ij}, \] so \(G_{ij}\in M_{n,n}^{\text{anti-sym}}(\mathbb{F})\). Further, it's not hard to show that \(\{G_{ij} : 1\le i\lt j\le n\}\) is a basis for \(M_{n,n}^{\text{anti-sym}}(\mathbb{F})\). Thus a matrix is anti-symmetric if and only if \(a_{ij}=-a_{ji}\), which means that the matrix can be written as \[ \begin{pmatrix} a_{11} & a_{12} & \cdots & a_{1n} \\ a_{21} & a_{22} & \cdots & a_{2n} \\ \vdots & & \ddots & \vdots \\ a_{n1} & a_{n2} & \cdots & a_{nn} \\ \end{pmatrix} = \sum_{i=1}^{n-1} \sum_{j=i+1}^n a_{ij}G_{ij}. \] (Notice in particular that diagonal entries satisfy \(a_{ii}=0\), since they equal their own negatives.) Hence \[ \dim M_{n,n}^{\text{anti-sym}}(\mathbb{F})= \frac{n^2-n}{2}. \]
(e) Given any matrix \(A\in M_{n,n}(\mathbb{F})\), write \(A\) as \[ A = \frac{1}{2}(A+A^*) + \frac{1}{2}(A-A^*). \] Then it's easy to check that \(\frac{1}{2}(A+A^*)\) is symmetric and \(\frac{1}{2}(A-A^*)\) is anti-symmetric, which shows the desired sum formula. In order to show that it is a direct sum, it's enough to show (by a result in the book) that the only matrix that is both symmetric and anti-symmetric is the zero matrix. But if \(A\) is both symmetric and anti-symmetric, then \[ A = A^* = -A,\quad\text{so}\quad 2A=0,\quad\text{so}\quad A=0. \]


Return to the Math 540 Homework Page

Return to the Math 540 Home Page