v2.1.3 页面样式微调，将masthead调矮，将标题字号改小

README.md

@@ -76,8 +76,8 @@ Live2D的后端暂时使用了[fghrsh](https://www.fghrsh.net/post/170.html)提
 > <p style="text-indent:2em">章北海感到父亲的灵魂从冥冥中降落到飞船上，与他融为一体，他按动了操作界面上那个最后的按钮，心中默念出那个他用尽一生的努力所追求的指令：</p><p style="text-indent:2em">“‘自然选择’，前进四！”</p>
 
 ## 历程
-### 2019-04-12 v2.1.3
-页面样式微调，将masthead调矮
+### 2019-04-29 v2.1.3
+页面样式微调，将masthead调矮，将标题字号改小
 ### 2019-03-20 v2.1.2
 修复sidebar展开时回到顶部的问题
 ### 2019-03-18 v2.1.1

_config.yml

@@ -95,12 +95,6 @@ head: # 删改前确定知道自己在做什么
         value: stylesheet
       - key: href
         value: /public/css/lanyon.css
-  - tag: link
-    var:
-      - key: rel
-        value: stylesheet
-      - key: href
-        value: //fonts.loli.net/css?family=PT+Serif:400,400italic,700%7CPT+Sans:400
   - tag: link
     var:
       - key: rel

_layouts/page.html

@@ -36,14 +36,14 @@
   <!-- Wrap is the content to shift when toggling the sidebar. We wrap the
          content to avoid any CSS collisions with our real content. -->
   <div class="wrap">
-    <h3 class="masthead masthead-title">
+    <h1 class="masthead masthead-title">
       <div class="container">
         {{ page.title }}
         <a href="{{ site.baseurl }}/">
           <small>{{ site.title }}</small>
         </a>
       </div>
-    </h3>
+    </h1>
     <div class="container content">
       {{ content }}
     </div>

_posts/2019-04-25-Codeforces Round #554 (Div. 2).md

@@ -1,5 +1,5 @@
 ---
-title: Codeforces Round \#554 (Div. 2)
+title: Codeforces Round #554 (Div. 2)
 categories:
   - ACM
   - 题解
@@ -234,3 +234,89 @@ int main()
 	printf("%d ", rk[g.e[g.p[n - 2]].second]);
 }
 ```
+# [Neko Rules the Catniverse (Small Version)](https://vjudge.net/problem/CodeForces-1152F1)
+状压DP，用$f[i][j][mask]$来表示正在考虑第i个，已经考虑了第$j$个，后$m$个的选择状态是二进制位集$mask$。
+
+这里可以滚动数组。
+```cpp
+#include <bits/stdc++.h>
+using namespace std;
+void add(int &a, long long b, int M = 1e9 + 7) { a = (a + b) % M; }
+int n, k, m;
+int main()
+{
+	scanf("%d%d%d", &n, &k, &m);
+	vector<vector<int>> f(k + 1, vector<int>(1 << m, 0));
+	f[0][0] = 1;
+	for (int i = 0; i < n; ++i)
+	{
+		vector<vector<int>> g(k + 1, vector<int>(1 << m, 0));
+		for (int j = 0; j < f.size(); ++j)
+			for (int mask = 0; mask < f[j].size(); ++mask)
+			{
+				int newMask = (mask << 1) % f[j].size();
+				add(g[j][newMask], f[j][mask]);
+				if (j < k)
+					add(g[j + 1][newMask | 1], (1LL + __builtin_popcount(mask)) * f[j][mask]);
+			}
+		f.swap(g);
+	}
+	for (int mask = m = 0; mask < f[k].size(); ++mask)
+		add(m, f[k][mask]);
+	printf("%d", m);
+}
+```
+# [Neko Rules the Catniverse (Large Version)](https://vjudge.net/problem/CodeForces-1152F2)
+注意到状态只在相邻的两位转移，这里可以用矩阵加速DP。
+
+下面这段代码跑了5070ms，官方还有一个[跑了156ms的解法](https://codeforces.com/contest/1152/submission/53260183)，暂时没看懂…也太强了吧。
+```cpp
+#include <bits/stdc++.h>
+using namespace std;
+int n, k, m;
+void add(int &a, long long b, int M = 1e9 + 7) { a = (a + b) % M; }
+int toId(int j, int mask) { return j << m | mask; }
+struct Matrix
+{
+	vector<vector<int>> a;
+	int n;
+	Matrix(int n) : n(n), a(n, vector<int>(n)) {}
+	friend Matrix operator*(const Matrix &a, const Matrix &b)
+	{
+		Matrix c(a.n);
+		for (int i = 0; i < a.n; ++i)
+			for (int j = 0; j < a.n; ++j)
+				for (int k = 0; k < a.n; ++k)
+					add(c.a[i][j], 1LL * a.a[i][k] * b.a[k][j]);
+		return c;
+	}
+	friend Matrix pow(Matrix a, int b)
+	{
+		Matrix r(a.n);
+		for (int i = 0; i < a.n; ++i)
+			r.a[i][i] = 1;
+		for (; b; b >>= 1, a = a * a)
+			if (b & 1)
+				r = a * r;
+		return r;
+	}
+};
+int main()
+{
+	scanf("%d%d%d", &n, &k, &m);
+	Matrix dp(toId(k + 1, 0)), f(dp);
+	for (int j = 0; j <= k; ++j)
+		for (int mask = 0; mask < (1 << m); ++mask)
+		{
+			int newMask = (mask << 1) % (1 << m);
+			f.a[toId(j, newMask)][toId(j, mask)] = 1;
+			if (j < k)
+				f.a[toId(j + 1, newMask | 1)][toId(j, mask)] = 1 + __builtin_popcount(mask);
+		}
+	dp.a[0][0] = 1, dp = pow(f, n) * dp;
+	int ans = 0;
+	for (int mask = 0; mask < (1 << m); ++mask)
+		add(ans, dp.a[toId(k, mask)][0]);
+	printf("%d", ans);
+}
+```