東京工業大学 情報理工学院 数理・計算科学系 2017年8月実施 午前 問C
Author
祭音Myyura
Description
以下の図に示すような n × m n \times m n × m の格子を考える.ただし n ≥ m ≥ 1 n \geq m \geq 1 n ≥ m ≥ 1 とする.
ここで,点 ( 0 , 0 ) (0, 0) ( 0 , 0 ) から点 ( n , m ) (n, m) ( n , m ) まで,条件 1 を満たすように隣接する点の間を移動する経路全体の集合を P ( n , m ) \mathcal{P}(n, m) P ( n , m ) とする.
条件 1: 経路上の全ての点に関して,その右または上の隣接点へのみ移動可能.
有限集合 A に対して |A| は A の濃度 (要素数) を表すものとする.
(1) ∣ P ( n , m ) ∣ |\mathcal{P}(n,m)| ∣ P ( n , m ) ∣ を求めよ.
P ( n , m ) \mathcal{P}(n,m) P ( n , m ) に含まれる経路で,条件 2 を満たす経路全体の集合を B ( n , m ) ⊆ P ( n , m ) \mathcal{B}(n, m) \subseteq \mathcal{P}(n, m) B ( n , m ) ⊆ P ( n , m ) とする.
条件 2: ある a ∈ { 1 , . . . , m } a \in \{1, . . . , m\} a ∈ { 1 , ... , m } に対して,経路上に点 ( a , a ) (a, a) ( a , a ) が存在する.すなわち,経路は上図に示した • 印の点を少なくとも 1 つ経由する.
集合 B ( n , m ) \mathcal{B}(n,m) B ( n , m ) を B ( 0 , 1 ) ( n , m ) ⊆ B ( n , m ) \mathcal{B}_{(0,1)}(n,m) \subseteq \mathcal{B}(n,m) B ( 0 , 1 ) ( n , m ) ⊆ B ( n , m ) と B ( 1 , 0 ) ⊆ B ( n , m ) \mathcal{B}_{(1,0)} \subseteq \mathcal{B}(n,m) B ( 1 , 0 ) ⊆ B ( n , m ) に分割する.
B ( 0 , 1 ) ( n , m ) : B ( n , m ) \mathcal{B}_{(0,1)}(n,m): \mathcal{B}(n,m) B ( 0 , 1 ) ( n , m ) : B ( n , m ) に含まれる経路で,点 ( 0 , 1 ) (0, 1) ( 0 , 1 ) を通る経路全体の集合.
B ( 1 , 0 ) ( n , m ) : B ( n , m ) \mathcal{B}_{(1,0)}(n,m): \mathcal{B}(n,m) B ( 1 , 0 ) ( n , m ) : B ( n , m ) に含まれる経路で,点 ( 1 , 0 ) (1, 0) ( 1 , 0 ) を通る経路全体の集合.
(2) ∣ B ( 0 , 1 ) ( n , m ) ∣ |\mathcal{B}_{(0,1)}(n,m)| ∣ B ( 0 , 1 ) ( n , m ) ∣ を求めよ.
(3) ∣ B ( 1 , 0 ) ( n , m ) ∣ |\mathcal{B}_{(1,0)}(n,m)| ∣ B ( 1 , 0 ) ( n , m ) ∣ を求めよ.
(4) r ( n , m ) = ∣ G ( n , m ) ∣ ∣ P ( n , m ) ∣ r(n,m) = \frac{|\mathcal{G}(n,m)|}{|\mathcal{P}(n,m)|} r ( n , m ) = ∣ P ( n , m ) ∣ ∣ G ( n , m ) ∣ | を求めよ.ただし G ( n , m ) = P ( n , m ) ∖ B ( n , m ) \mathcal{G}(n,m) = \mathcal{P}(n,m) \setminus \mathcal{B}(n,m) G ( n , m ) = P ( n , m ) ∖ B ( n , m ) .
また n = 15 , m = 10 n = 15, m = 10 n = 15 , m = 10 として r ( 15 , 10 ) r(15, 10) r ( 15 , 10 ) を求めよ.
题目描述
考虑题面图示的 n × m n\times m n × m 方格,其中 n ≥ m ≥ 1 n\geq m\geq1 n ≥ m ≥ 1 。令 P ( n , m ) \mathcal P(n,m) P ( n , m ) 为所有从 ( 0 , 0 ) (0,0) ( 0 , 0 ) 走到 ( n , m ) (n,m) ( n , m ) 且满足条件 1 的格路径集合。
条件 1: 在路径上的每一点,只能移动到其右侧或上侧的相邻点。
对有限集合 A A A ,以 ∣ A ∣ |A| ∣ A ∣ 表示其元素个数。
求 ∣ P ( n , m ) ∣ |\mathcal P(n,m)| ∣ P ( n , m ) ∣ 。
在 P ( n , m ) \mathcal P(n,m) P ( n , m ) 中,令 B ( n , m ) \mathcal B(n,m) B ( n , m ) 为满足条件 2 的路径集合。
条件 2: 存在某个 a ∈ { 1 , … , m } a\in\{1,\ldots,m\} a ∈ { 1 , … , m } ,使路径经过点 ( a , a ) (a,a) ( a , a ) ;也就是说,路径至少经过题面图中一个以黑点标出的正对角点。
再把 B ( n , m ) \mathcal B(n,m) B ( n , m ) 分为以下两个子集:
B ( 0 , 1 ) ( n , m ) \mathcal B_{(0,1)}(n,m) B ( 0 , 1 ) ( n , m ) :B ( n , m ) \mathcal B(n,m) B ( n , m ) 中经过点 ( 0 , 1 ) (0,1) ( 0 , 1 ) 的全部路径;
B ( 1 , 0 ) ( n , m ) \mathcal B_{(1,0)}(n,m) B ( 1 , 0 ) ( n , m ) :B ( n , m ) \mathcal B(n,m) B ( n , m ) 中经过点 ( 1 , 0 ) (1,0) ( 1 , 0 ) 的全部路径。
求 ∣ B ( 0 , 1 ) ( n , m ) ∣ |\mathcal B_{(0,1)}(n,m)| ∣ B ( 0 , 1 ) ( n , m ) ∣ 。
求 ∣ B ( 1 , 0 ) ( n , m ) ∣ |\mathcal B_{(1,0)}(n,m)| ∣ B ( 1 , 0 ) ( n , m ) ∣ 。
定义
G ( n , m ) = P ( n , m ) ∖ B ( n , m ) , r ( n , m ) = ∣ G ( n , m ) ∣ ∣ P ( n , m ) ∣ . \mathcal G(n,m)=\mathcal P(n,m)\setminus\mathcal B(n,m),
\qquad
r(n,m)=\frac{|\mathcal G(n,m)|}{|\mathcal P(n,m)|}. G ( n , m ) = P ( n , m ) ∖ B ( n , m ) , r ( n , m ) = ∣ P ( n , m ) ∣ ∣ G ( n , m ) ∣ .
求 r ( n , m ) r(n,m) r ( n , m ) 的表达式,并进一步计算 n = 15 , m = 10 n=15,m=10 n = 15 , m = 10 时的 r ( 15 , 10 ) r(15,10) r ( 15 , 10 ) 。
格路径与二项式系数:把向右、向上的步序列计数转化为组合数。
对角线首次相交的路径计数:利用路径分割或反射对应关系,分别计算经 ( 0 , 1 ) (0,1) ( 0 , 1 ) 、( 1 , 0 ) (1,0) ( 1 , 0 ) 且触及正对角点的路径。
补集概率:由总路径数减去经过指定对角点的路径数,化简合法路径比例并代入给定参数。
Kai
(1)
Total number of moves in which we have to move up to reach the last row = m = m = m
Total number of moves in which we have to move right to reach the last column = n = n = n
Total moves = m + n = m + n = m + n
Now think of moves as a string of ‘R’ and ‘U’ characters where ‘R’ at any ith index will tell us to move ‘Right’ and ‘U’ will tell us to move ‘Up’.
Then P ( n , m ) \mathcal{P}(n,m) P ( n , m ) equals to how many unique strings (moves) we can make where in total there should be ( n + m ) (n + m) ( n + m ) characters and there should be m m m ‘U’ character and n n n ‘R’ character.
Note that choosing positions of n n n ‘R’ characters results in the automatic choosing of m m m ‘U’ character positions
Therefore,
P ( n , m ) = ( n + m m ) \mathcal{P}(n,m) = \binom{n+m}{m} P ( n , m ) = ( m n + m )
(2)
Note that all paths go through the point ( 0 , 1 ) (0,1) ( 0 , 1 ) will cross the line y = x y=x y = x , hence
B ( 0 , 1 ) ( n , m ) = P ( n , m − 1 ) = ( n + m − 1 m − 1 ) \mathcal{B}_{(0,1)} (n,m) = \mathcal{P}(n, m-1) = \binom{n+m-1}{m-1} B ( 0 , 1 ) ( n , m ) = P ( n , m − 1 ) = ( m − 1 n + m − 1 )
(3)
We take a path from ( 1 , 0 ) (1,0) ( 1 , 0 ) to ( n , m ) (n,m) ( n , m ) which touch (cross) the line y = x y=x y = x ,
and let p p p be the point where it first touches this line.
If we reflect the portion of the path from ( 1 , 0 ) (1,0) ( 1 , 0 ) to p p p around the line y = x y=x y = x , then we get a path from ( 0 , 1 ) (0,1) ( 0 , 1 ) to ( n , m ) (n,m) ( n , m ) .
Conversely, any path from ( 0 , 1 ) (0,1) ( 0 , 1 ) to ( n , m ) (n,m) ( n , m ) must cross the line y = x y=x y = x .
If we let p ′ p' p ′ be the point where the path first touches this line and reflect the portion of the path from ( 0 , 1 ) (0,1) ( 0 , 1 ) to p ′ p' p ′ about this line, then we get a path from ( 1 , 0 ) (1,0) ( 1 , 0 ) to ( n , m ) (n,m) ( n , m ) .
Since this gives a bijection between the two types of paths, and since the number of paths from ( 0 , 1 ) (0,1) ( 0 , 1 ) to ( n , m ) (n,m) ( n , m ) is given by B ( 0 , 1 ) ( n , m ) \mathcal{B}_{(0,1)} (n,m) B ( 0 , 1 ) ( n , m ) , we have
B ( 1 , 0 ) ( n , m ) = B ( 0 , 1 ) ( n , m ) = ( n + m − 1 m − 1 ) \mathcal{B}_{(1,0)} (n,m) = \mathcal{B}_{(0,1)} (n,m) = \binom{n+m-1}{m-1} B ( 1 , 0 ) ( n , m ) = B ( 0 , 1 ) ( n , m ) = ( m − 1 n + m − 1 )
(4)
By (2) and (3) we have
r ( n , m ) = ∣ G ( n , m ) ∣ ∣ P ( n , m ) ∣ = ( n + m m ) − 2 ( n + m − 1 m − 1 ) ( n + m m ) = 1 − 2 ( n + m − 1 m − 1 ) ( n + m m ) = 1 − 2 ⋅ ( n + m − 1 ) ! ( m − 1 ) ! n ! ⋅ m ! n ! ( m + n ) ! = 1 − 2 m m + n \begin{aligned}
r(n,m) &= \frac{|\mathcal{G}(n,m)|}{|\mathcal{P}(n,m)|} = \frac{\binom{n+m}{m}-2\binom{n+m-1}{m-1}}{\binom{n+m}{m}} \\
&= 1 - \frac{2\binom{n+m-1}{m-1}}{\binom{n+m}{m}} \\
&= 1 - 2 \cdot \frac{(n+m-1)!}{(m-1)!n!} \cdot \frac{m!n!}{(m+n)!} \\
&= 1 - \frac{2m}{m+n}
\end{aligned} r ( n , m ) = ∣ P ( n , m ) ∣ ∣ G ( n , m ) ∣ = ( m n + m ) ( m n + m ) − 2 ( m − 1 n + m − 1 ) = 1 − ( m n + m ) 2 ( m − 1 n + m − 1 ) = 1 − 2 ⋅ ( m − 1 )! n ! ( n + m − 1 )! ⋅ ( m + n )! m ! n ! = 1 − m + n 2 m
Then
r ( 15 , 10 ) = 1 − 2 × 10 10 + 15 = 1 5 r(15, 10) = 1 - \frac{2 \times10}{10 + 15} = \frac{1}{5} r ( 15 , 10 ) = 1 − 10 + 15 2 × 10 = 5 1