changed 3 years ago
Linked with GitHub

西爾維斯特矩陣、結式

Creative Commons License
This work by Jephian Lin is licensed under a Creative Commons Attribution 4.0 International License.

\(\newcommand{\trans}{^\top} \newcommand{\adj}{^{\rm adj}} \newcommand{\cof}{^{\rm cof}} \newcommand{\inp}[2]{\left\langle#1,#2\right\rangle} \newcommand{\dunion}{\mathbin{\dot\cup}} \newcommand{\bzero}{\mathbf{0}} \newcommand{\bone}{\mathbf{1}} \newcommand{\ba}{\mathbf{a}} \newcommand{\bb}{\mathbf{b}} \newcommand{\bc}{\mathbf{c}} \newcommand{\bd}{\mathbf{d}} \newcommand{\be}{\mathbf{e}} \newcommand{\bh}{\mathbf{h}} \newcommand{\bp}{\mathbf{p}} \newcommand{\bq}{\mathbf{q}} \newcommand{\br}{\mathbf{r}} \newcommand{\bx}{\mathbf{x}} \newcommand{\by}{\mathbf{y}} \newcommand{\bz}{\mathbf{z}} \newcommand{\bu}{\mathbf{u}} \newcommand{\bv}{\mathbf{v}} \newcommand{\bw}{\mathbf{w}} \newcommand{\tr}{\operatorname{tr}} \newcommand{\nul}{\operatorname{null}} \newcommand{\rank}{\operatorname{rank}} %\newcommand{\ker}{\operatorname{ker}} \newcommand{\range}{\operatorname{range}} \newcommand{\Col}{\operatorname{Col}} \newcommand{\Row}{\operatorname{Row}} \newcommand{\spec}{\operatorname{spec}} \newcommand{\vspan}{\operatorname{span}} \newcommand{\Vol}{\operatorname{Vol}} \newcommand{\sgn}{\operatorname{sgn}} \newcommand{\idmap}{\operatorname{id}} \newcommand{\am}{\operatorname{am}} \newcommand{\gm}{\operatorname{gm}} \newcommand{\mult}{\operatorname{mult}} \newcommand{\iner}{\operatorname{iner}}\)

from lingeo import random_int_list
from linspace import vtop, ptov, syl

Main idea

Let \(p\) and \(q\) be two polynomials.
The greatest common divisor of \(p\) and \(q\) is the polynomial \(g\) of largest degree such that \(g \mid p\) and \(g\mid q\).
This polynomial is unique up to scalar multiplication, so usually we let \(\gcd(p,q)\) be the one with leading coefficeint \(1\).

By the Euclidean algorithm, it is known that the following two sets are equal.
\[\{ap + bq: a,b\in \mathcal{P}\} = \{ag: a\in\mathcal{P}\}, \] where \(=\mathcal{P}\) is the set of all polynomials.

A refined version is as follows.
Let \(p\) and \(q\) be two polynomials of degree \(m\) and \(n\), respectively.
Then
\[\{ap + bq \in\mathcal{P}_{m+n}: a\in\mathcal{P}_{n-1},b\in\mathcal{P}_{m-1}\} = \{ag \in\mathcal{P}_{m+n}: a\in\mathcal{P}\}. \]

Given two polynomials \(p, q\) of degrees \(m,n\), consider the function
\[\begin{array}{rccc} f : & \mathcal{P}_{n-1} \times \mathcal{P}_{m-1} & \rightarrow & \mathcal{P}_{m+n-1} \\ & (a,b) & \mapsto & ap + bq \\ \end{array}, \] which is linear.
Thus,
\[\operatorname{range}(f) = \{ap + bq \in\mathcal{P}_{m+n-1}: a\in\mathcal{P}_{n-1},b\in\mathcal{P}_{m-1}\} = \{ag \in\mathcal{P}_{m+n-1}: a\in\mathcal{P}\}, \] where \(g = \gcd(p,q)\).
Therefore, \(f\) is surjective if and only if \(\gcd(p,q) = 1\).

Let \(\alpha_q = \{1,\ldots, x^{n-1}\}\) and \(\alpha_q = \{1,\ldots, x^{m-1}\}\) be the standard bases of \(\mathcal{P}_{n-1}\) and \(\mathcal{P}_{m-1}\).
Let \[\alpha = \{(a,0): a\in\alpha_p\} \cup \{(0,b) : b\in\alpha_q\}. \]
Then \(\alpha\) is a basis of \(\mathcal{P}_{n-1}\times\mathcal{P}_{m-1}\).
On the other hand, let \(\beta\) be the standard basis of \(\mathcal{P}_{m+n-1}\).

Construct the \((m + n)\times (m + n)\) matrix
\[S_{p,q} = \begin{bmatrix} | & ~ & | & | & ~ & | \\ [p]_\beta & \cdots & [x^{n-1}p]_\beta & [q]_\beta & \cdots & [x^{m-1}q]_\beta \\ | & ~ & | & | & ~ & | \\ \end{bmatrix}. \]
Then \(S_{p,q} = [f]_\alpha^\beta\) and is called the Sylvester matrix of \(p\) and \(q\).
The determinant of \(S_{p,q}\) is called the resultant of \(p\) and \(q\), denoted as \(\operatorname{res}(p,q)\).
(We have not learnt the properties of the determinant, but at least it make senses when \(S_{p,q}\) is a small matrix.)

Let \(p,q\) be two polynomials.
Let \(S_{p,q}\) their Sylvester matrix and \(\operatorname{res}(p,q)\) the resultant.
Then the following are equivalent:

  1. \(\gcd(p,q) = 1\).
  2. \(f\) is surjective.
  3. \(f\) is injective.
  4. \(S_{p,q}\) is invertible.
  5. \(\operatorname{res}(p,q)\neq 0\).

Side stories

  • gcd by row operations
  • multiple root

Experiments

Exercise 1

執行以下程式碼。

### code
set_random_seed(0)
print_ans = False
m,n = 2,3
p = vtop(vector(random_int_list(m+1)))
q = vtop(vector(random_int_list(n+1)))

print("p =", p)
print("q =", q)

A = syl(p,q)

if print_ans:
    print("alpha = {(1,0), (x,0), (x^2,0), (0,1), (0,x)}")
    print("beta = {1, x, x^2, x^3, x^4}")
    print("Spq =")
    show(A)
    print("gcd(p,q) = 1?", (A.determinant() != 0))

set_random_seed(0)

\(p= 5x^2 + 3x - 4\)

\(q=3x^3 - 5x - 5\)

\(\alpha = \{(1,0), (x,0), (x^2,0), (0,1), (0,x)\}\)

\(\beta = \{1, x, x^2, x^3, x^4\}\)

\[S_{p,q}=\begin{bmatrix} -4&0&0&-5&0\\ 3&-4&0&-5&-5\\ 5&3&-4&0&-5\\ 0&5&3&3&0\\ 0&0&5&0&3 \end{bmatrix} \]

Exercise 1(a)

寫出 \(\mathcal{P}_2\times\mathcal{P}_1\) 的標準基底 \(\alpha\)
以及 \(\mathcal{P}_4\)

  • 標點

\(Ans\)

\(\mathcal{P}_2\) 的標準基底 \(\alpha_2=\{1, x, x^2\}\)

\(\mathcal{P}_1\) 的標準基底 \(\alpha_1=\{1,x\}\)

\(\mathcal{P}_2\times\mathcal{P}_1\) 的標準基底 \(\alpha = \{(a,0): a\in\alpha_2\} \cup \{(0,b) : b\in\alpha_1\}\)

得知 \(\alpha= \{(1,0), (x,0), (x^2,0), (0,1), (0,x)\}\)

\(\mathcal{P}_4\) 的標準基底為 \(\{1,x,x^2,x^3,x^4\}\)

Exercise 1(b)

寫出 \(p,q\) 的西爾維斯特矩陣 \(A\)

  • 標點
  • 句子要完整

\(Ans\)

西爾維斯特矩陣為 \[S_{p,q} = \begin{bmatrix} | & ~ & | & | & ~ & | \\ [p]_\beta & \cdots & [x^{n-1}p]_\beta & [q]_\beta & \cdots & [x^{m-1}q]_\beta \\ | & ~ & | & | & ~ & | \\ \end{bmatrix}. \]
\(n=3\) , \(m=2\) , \(p= 3x^3 - 5x - 5\) , \(q=5x^2 + 3x - 4\) 代入,

得到 \(p,q\) 的西爾維斯特矩陣

\(A=\begin{bmatrix} -4&0&0&-5&0\\ 3&-4&0&-5&-5\\ 5&3&-4&0&-5\\ 0&5&3&3&0\\ 0&0&5&0&3 \end{bmatrix}\)

Exercise 1©

判斷 \(p,q\) 是否互質。

  • 標點

\(Ans\)

是,\(S_{p,q}\) 經過列運算之後,發現沒有自由變數,表示此矩陣可逆,也就代表 \(\gcd(p,q) = 1\)

\(p,q\) 互質。

Exercises

Exercise 2

對以下的 \(p\)\(q\)﹐利用西爾維斯特矩陣判斷它們是否互質。

Exercise 2(a)

\(p = 1 + 2x + x^2\)
\(q = 2 + x\)

  • 中英數間空格
  • \(det\) > \(\det\)
  • 後面幾小題一樣

Ans \[ S_{p,q}=\begin{bmatrix} 1&2&0\\ 2&1&2\\ 1&0&1 \end{bmatrix} \] 由於 \(\det(S_{p,q})=1\neq 0\),說明了 \(S_{p,q}\) 為一可逆矩陣,也代表 \(\gcd(p,q) = 1\),因此多項式 \(p\)\(q\) 互質。

Exercise 2(b)

\(p = 1 + 2x + x^2\)
\(q = 2 + 3x + x^2\)

ANS

\[ S_{p,q} = \begin{bmatrix} 1&0&2&0\\ 2&1&3&2\\ 1&2&1&3\\ 0&1&0&1 \end{bmatrix} \]

針對高階的矩陣,我們可以利用拉普拉斯展開法降階以獲取行列式值。
由於 \(\det(S_{p,q})=0\),說明了 \(S_{p,q}\) 為一不可逆矩陣,也代表 \(\gcd(p,q)\neq 1\),因此多項式 \(p\)\(q\) 不是互質。

Exercise 2©

\(p = 1 + 2x + x^2\)
\(q = 6 + 11x + 6x^2 + x^3\)

ANS

\[ S_{p,q}=\begin{bmatrix} 1&0&0&6&0\\ 2&1&0&11&6\\ 1&2&1&6&11\\ 0&1&2&1&6\\ 0&0&1&0&1 \end{bmatrix} \] 由於 \(\det(S_{p,q})=0\),說明了 \(S_{p,q}\) 為一不可逆矩陣,也代表 \(\gcd(p,q)\neq 1\) ,因此多項式 \(p\)\(q\) 不是互質。

Exercise 3

說明西爾維斯特矩陣 \(S_{p,q}\) 就是 \([f]_\alpha^\beta\)

good

\[\begin{array}{rccc} f : & \mathcal{P}_{n-1} \times \mathcal{P}_{m-1} & \rightarrow & \mathcal{P}_{m+n-1} \\ & (a,b) & \mapsto & ap + bq \\ \end{array}, \] 又令 \(\alpha\)\(\mathcal{P}_{n-1} \times \mathcal{P}_{m-1}\) 的基底,而 \(\beta\)\(\mathcal{P}_{m+n-1}\) 的基底。
\(\alpha = \{(1,0),(x,0),(x^2,0),...,(x^{n-1},0),(0,1),(0,x),(0,x^2),...,(0,x^{m-1})\}\)
接著觀察 \([f]_\alpha^\beta\)
依照定義, \([f]_\alpha^\beta\) 為將 \(\alpha\) 內的向量一一代入 \(f\) 並用 \(\beta\) 表示出來的矩陣,即 \[ [f]_\alpha^\beta = \begin{bmatrix} | & | & | & ~ & | & | & ~ & | \\ [f(1,0)]_\beta & [f(x,0)]_\beta & [f(x^2,0)]_\beta & ... & [f(x^{n-1},0)]_\beta & [f(0,1)]_\beta & ... & [f(0,x^{m-1})]_\beta \\ | & | & | & ~& | & | & ~ & | \end{bmatrix}。 \] 而觀察將 \(\alpha\) 內的向量代入 \(f\) 的結果,會發現 \(f(1,0) = p,f(x,0) = px,...,f(x^{n-1},0) = px^{n-1},f(0,1) = q,f(0,x^{m-1} = qx^{m-1})\)
\[ [f]_\alpha^\beta = \begin{bmatrix} | & ~ & | & | & ~ & | \\ [p]_\beta & \cdots & [x^{n-1}p]_\beta & [q]_\beta & \cdots & [x^{m-1}q]_\beta \\ | & ~ & | & | & ~ & | \\ \end{bmatrix}$ 即 $S_{p,q}. \]

Exercise 4

執行以下程式碼。
嘗試各種不同的 \(p,q\)
\(A\) 為它們的西爾維斯特矩陣﹐
\(A\) 左上和右下翻轉後得到 \(B\)
\(R\)\(B\) 的最簡階梯形式矩陣。
觀察 \(\gcd(p,q)\)\(R\) 的關係﹐並說明為什麼。

p = 1 + x
q = 1 + 2*x + x**2
A = syl(p,q)
B = A.antitranspose()
R = B.rref()
print("A =")
show(A)
print("B =")
show(B)
print("R =")
show(R)
print("gcd =", p.polynomial(QQ).gcd(q.polynomial(QQ)))
  • \(gcd\) > \(\gcd\)
  • 中英數間空格
  • 填上係數後可以發現皆有 \(p,q\) 的公因式存在 > 填上係數後可以發現皆是 \(\gcd(p,q)\) 的倍式
  • 且在矩陣\(R\)中,因為是由矩陣\(B\)所推倒而來,前幾列為 \(p,q\) 被整理過後的一元二次式,因此也會有相同的公因式。 > 且在矩陣 \(R\) 中,任一列都是 \(B\) 中列向量的線性組合,其所對應到的多項式皆為 \(ap + bq\) 的形式,所以均是 \(\gcd(p,q)\) 的倍式。而最後一個非零列的次方數最小,所以它就是 \(\gcd(p,q)\)

答:

\(p = 1 + x\), \(q = 1 + 2x + x^2\) \[A=\begin{bmatrix} 1&0&1\\ 1&1&2\\ 0&1&1\\ \end{bmatrix},B=\begin{bmatrix} 1&2&1\\ 1&1&0\\ 0&1&1\\ \end{bmatrix},R=\begin{bmatrix} 1&0&-1\\ 0&1&1\\ 0&0&0\\ \end{bmatrix},\gcd(p,q) = x+1 .\]

\(p = 1 + x^2\), \(q = 1 + 2x\) \[A=\begin{bmatrix} 1&1&0\\ 0&2&1\\ 1&0&2\\ \end{bmatrix},B=\begin{bmatrix} 2&1&0\\ 0&2&1\\ 1&0&1\\ \end{bmatrix},R=\begin{bmatrix} 1&0&0\\ 0&1&0\\ 0&0&1\\ \end{bmatrix},\gcd(p,q) = 1 .\]

\(p = 2 + x\), \(q = 4 + 8x + 3x^2\) \[A=\begin{bmatrix} 2&0&4\\ 1&2&8\\ 0&1&3\\ \end{bmatrix},B=\begin{bmatrix} 3&8&4\\ 1&2&0\\ 0&1&2\\ \end{bmatrix},R=\begin{bmatrix} 1&0&-4\\ 0&1&2\\ 0&0&0\\ \end{bmatrix},\gcd(p,q) = x+2 . \]

矩陣最下方不為零的列,按照順序從高次到低次排序,分別是 \(x^2, x^1, x^0\),填上係數就是 \(p,q\) 的公因式。其中在矩陣 \(R\) 的第一列由上往下到不為零的前一列,填上係數後可以發現皆是 \(\gcd(p,q)\) 的倍式,且在矩陣 \(R\) 中,任一列都是 \(B\) 中列向量的線性組合,其所對應到的多項式皆為 \(ap + bq\) 的形式,所以均是 \(\gcd(p,q)\) 的倍式。而最後一個非零列的次方數最小,所以它就是 \(\gcd(p,q)\)

Exercise 5

\(p,q\) 為兩多項式且 \(g = \gcd(p,q)\)
證明
\[\{ap + bq: a,b\in \mathcal{P}\} = \{ag: a\in\mathcal{P}\}. \]

一邊是簡單的,\(ap + bp\) 一定是 \(cg\) 的形式。
另一邊要用到除法原理。
這題我直接掛懸賞。

ANS
因為 \(g\)\(p,q\) 公因式,因此我們可令 \({\{ap=g: a\in\mathcal{P}\}}\)\({\{bq=g: d\in\mathcal{P}\}}\) ,不會了

目前分數 6.5

Select a repo