Learn Linear Algebra

Theorem

The column space of an m×n m \times n matrix A A is all of Rm \mathbb{R}^m if and only if the equation Ax=b A\vec{x} = \vec{b} has a solution for each bRm \vec{b} \in \mathbb{R}^m .

Proof:

We want to prove that the column space of an m×n m \times n matrix A A is all of Rm \mathbb{R}^m if and only if the equation Ax=b A \vec{x} = \vec{b} has a solution for every bRm \vec{b} \in \mathbb{R}^m .

Col(A)=RmFor every bRm,b can be expressed as a linear combination of the columns of A.For every bRm,bCol(A)For every bRm,xRn such that Ax=b. \text{Col}(A) = \mathbb{R}^m \\ \Leftrightarrow \text{For every } \vec{b} \in \mathbb{R}^m, \vec{b} \text{ can be expressed as a linear combination of the columns of } A. \\ \Leftrightarrow \text{For every } \vec{b} \in \mathbb{R}^m, \vec{b} \in \text{Col}(A) \\ \Leftrightarrow \text{For every } \vec{b} \in \mathbb{R}^m, \exists \, \vec{x} \in \mathbb{R}^n \text{ such that } A \vec{x} = \vec{b}.

Thus, the column space of A A is all of Rm \mathbb{R}^m if and only if the equation Ax=b A \vec{x} = \vec{b} has a solution for every bRm \vec{b} \in \mathbb{R}^m .

Theorem

The null space of an m×n m \times n matrix A A is a subspace of Rn \mathbb{R}^n . Equivalently, the set of all solutions to a system Ax=0 A\vec{x} = \vec{0} of m m homogeneous linear equations in n n unknowns is a subspace of Rn \mathbb{R}^n .

Proof:

Let A A be an m×n m \times n matrix. The null space of A A , denoted as Nul(A) \text{Nul}(A) , is defined as: Nul(A)={xRnAx=0}. \text{Nul}(A) = \{ \vec{x} \in \mathbb{R}^n \mid A \vec{x} = \vec{0} \}. We will show that Nul(A) \text{Nul}(A) is a subspace of Rn \mathbb{R}^n by verifying the three subspace properties:

1. The zero vector is in Nul(A) \text{Nul}(A) :
Let x=0 \vec{x} = \vec{0} . Then: A0=0. A \vec{0} = \vec{0}. Hence, 0Nul(A) \vec{0} \in \text{Nul}(A) .

2. Closed under addition:
Let u,wNul(A) \vec{u}, \vec{w} \in \text{Nul}(A) . Then: Au=0andAw=0. A \vec{u} = \vec{0} \quad \text{and} \quad A \vec{w} = \vec{0}. Adding these equations gives: A(u+w)=Au+Aw=0+0=0. A (\vec{u} + \vec{w}) = A \vec{u} + A \vec{w} = \vec{0} + \vec{0} = \vec{0}. Hence, u+wNul(A) \vec{u} + \vec{w} \in \text{Nul}(A) .

3. Closed under scalar multiplication:
Let uNul(A) \vec{u} \in \text{Nul}(A) and let cR c \in \mathbb{R} . Then: Au=0. A \vec{u} = \vec{0}. Multiplying by c c gives: A(cu)=c(Au)=c0=0. A (c \vec{u}) = c (A \vec{u}) = c \vec{0} = \vec{0}. Hence, cuNul(A) c \vec{u} \in \text{Nul}(A) .

Thus, Nul(A) \text{Nul}(A) satisfies the conditions for a subspace of Rn \mathbb{R}^n .

Theorem

The column space of an m×n m \times n matrix A A is a subspace of Rm \mathbb{R}^m .

Proof:

Let A A be an m×n m \times n matrix. The column space of A A , denoted by Col(A) \text{Col}(A) , is defined as: Col(A)={AxxRn}. \text{Col}(A) = \{ A \vec{x} \mid \vec{x} \in \mathbb{R}^n \}. This is the set of all linear combinations of the columns of A A . We will show that Col(A) \text{Col}(A) is a subspace of Rm \mathbb{R}^m by verifying the three subspace properties:

1. The zero vector is in Col(A) \text{Col}(A) :
Let x=0Rn \vec{x} = \vec{0} \in \mathbb{R}^n . Then: Ax=A0=0. A \vec{x} = A \vec{0} = \vec{0}. Hence, 0Col(A) \vec{0} \in \text{Col}(A) .

2. Closed under addition:
Let u,wCol(A) \vec{u}, \vec{w} \in \text{Col}(A) . Then there exist x1,x2Rn \vec{x}_1, \vec{x}_2 \in \mathbb{R}^n such that: u=Ax1andw=Ax2. \vec{u} = A \vec{x}_1 \quad \text{and} \quad \vec{w} = A \vec{x}_2. Adding u \vec{u} and w \vec{w} gives: u+w=Ax1+Ax2=A(x1+x2). \vec{u} + \vec{w} = A \vec{x}_1 + A \vec{x}_2 = A (\vec{x}_1 + \vec{x}_2). Since x1+x2Rn \vec{x}_1 + \vec{x}_2 \in \mathbb{R}^n , u+wCol(A) \vec{u} + \vec{w} \in \text{Col}(A) .

3. Closed under scalar multiplication:
Let uCol(A) \vec{u} \in \text{Col}(A) and cR c \in \mathbb{R} . Then there exists xRn \vec{x} \in \mathbb{R}^n such that: u=Ax. \vec{u} = A \vec{x}. Multiplying by c c gives: cu=c(Ax)=A(cx). c \vec{u} = c (A \vec{x}) = A (c \vec{x}). Since cxRn c \vec{x} \in \mathbb{R}^n , cuCol(A) c \vec{u} \in \text{Col}(A) .

Thus, Col(A) \text{Col}(A) satisfies the conditions for a subspace of Rm \mathbb{R}^m .