Chapter 4 (Vector Spaces): Linearly independent sets； Bases (基)

本文为《Linear algebra and its applications》的读书笔记

• In this section we identify and study the subsets that span a vector space $V$ or a subspace $H$ as “efficiently” as possible. The key idea is that of linear independence.

Linearly independent sets

• The definition of linear independence in a vector space is just the same as in ${\mathbb{R}}^n$.

• The main difference between linear dependence in ${\mathbb{R}}^n$ and in a general vector space is that when the vectors are not $n$-tuples, the homogeneous equation
  usually cannot be written as a system of $n$ linear equations. That is, the vectors cannot be made into the columns of a matrix $A$ in order to study the equation $A\mathbf{x}=0$. We must rely instead on the definition of linear dependence and on Theorem 4.

---

EXAMPLE 1

• Let ${\mathbf{p}}_1(t)=1,{\mathbf{p}}_2(t)=t$ , and ${\mathbf{p}}_3(t)=4-t$. Then { p 1 , p 2 , p 3 } \{\boldsymbol p_1,\boldsymbol p_2,\boldsymbol p_3\} {p1​,p2​,p3​} is linearly dependent in $\mathbb{P}$ because ${\mathbf{p}}_3=4{\mathbf{p}}_1-{\mathbf{p}}_2$.

基 / 极大线性无关组 / 极小生成集

Observe that when $H\ne V$ , condition $(ii)$ includes the requirement that each of the vectors { b 1 , . . . , b p } \{\boldsymbol b_1,...,\boldsymbol b_p\} {b1​,...,bp​} must belong to $H$

---

EXAMPLE 3

• Let $A$ be an invertible $n\times n$ matrix. Then the columns of $A$ form a basis for ${\mathbb{R}}^n$.

---

EXAMPLE 4

• Let { e 1 , . . . , e n } \{\boldsymbol e_1,...,\boldsymbol e_n\} {e1​,...,en​} be the columns of the $n\times n$ identity matrix, $I_n$. The set { e 1 , . . . , e n } \{\boldsymbol e_1,...,\boldsymbol e_n\} {e1​,...,en​} is called the standard basis (标准基) for ${\mathbb{R}}^n$ (Figure 1).
  

---

EXAMPLE 6

• Let S = { 1 , t , t 2 , . . . , t n } S = \{1, t, t^2,..., t^n\} S={1,t,t2,...,tn}. $S$ is is called the standard basis for ${\mathbb{P}}^n$.
  

The Spanning Set Theorem

生成集定理

• As we will see, a basis is an “efficient” spanning set that contains no unnecessary vectors.
• In fact, a basis can be constructed from a spanning set by discarding unneeded vectors.

Two Views of a Basis

• When the Spanning Set Theorem is used, the deletion of vectors from a spanning set must stop when the set becomes linearly independent. Thus a basis is a spanning set that is as small as possible.
• A basis is also a linearly independent set that is as large as possible. If $S$ is a basis for $V$, and if $S$ is enlarged by one vector—say, $\mathbf{w}$—from $V$ , then the new set cannot be linearly independent, because $S$ spans $V$ , and $\mathbf{w}$ is therefore a linear combination of the elements in $S$.

Bases for $NulA$ and $ColA$

• We already know how to find vectors that span the null space of a matrix $A$. The discussion in Section 4.2 pointed out that our method always produces a linearly independent set when $NulA$ contains nonzero vectors. So, in this case, that method produces a basis for $NulA$.
• The next two examples describe a simple algorithm for finding a basis for the column space.

---

EXAMPLE 8

• Find a basis for $ColB$, where
  

SOLUTION

• Each nonpivot column of $B$ is a linear combination of the pivot columns. By the Spanning Set Theorem, we may discard ${\mathbf{b}}_2$ and ${\mathbf{b}}_4$, and S = { b 1 , b 3 , b 5 } S=\{\boldsymbol b_1,\boldsymbol b_3,\boldsymbol b_5\} S={b1​,b3​,b5​} will still span $ColB$. $S$ is a basis for $ColB$.

---

• What about a matrix $A$ that is not in reduced echelon form? Recall that any linear dependence relationship among the columns of $A$ can be expressed in the form $A\mathbf{x}=0$, where $\mathbf{x}$ is a column of weights. When $A$ is row reduced to a matrix $B$, the columns of $B$ are often totally different from the columns of $A$. However, the equations $A\mathbf{x}=0$ and $B\mathbf{x}=0$ have exactly the same set of solutions. Then the vector equations
  also have the same set of solutions.
• That is, the columns of $A$ have exactly the same linear dependence relationships as the columns of $B$.

---

EXAMPLE 9

It can be shown that the matrix
is row equivalent to the matrix $B$ in Example 8. Find a basis for $ColA$.

SOLUTION

• In Example 8 we saw that
  so we can expect that
  Thus { a 1 , a 3 , a 5 } \{\boldsymbol a_1,\boldsymbol a_3,\boldsymbol a_5\} {a1​,a3​,a5​} is a basis for $ColA$. The columns we have used for this basis are the pivot columns of $A$.