This commit is contained in:
krahets
2023-11-09 05:13:54 +08:00
parent 9a09f9407e
commit 3f666fa676
85 changed files with 619 additions and 610 deletions

View File

@@ -3415,7 +3415,7 @@ G & = \{ V, E \} \newline
\end{aligned}
\]</div>
<p>如果将顶点看作节点,将边看作连接各个节点的引用(指针),我们就可以将图看作是一种从链表拓展而来的数据结构。如图 9-1 所示,<strong>相较于线性关系(链表)和分治关系(树),网络关系(图)的自由度更高</strong>,从而更为复杂。</p>
<p><a class="glightbox" href="../graph.assets/linkedlist_tree_graph.png" data-type="image" data-width="100%" data-height="auto" data-desc-position="bottom"><img alt="链表、树、图之间的关系" src="../graph.assets/linkedlist_tree_graph.png" /></a></p>
<p><a class="glightbox" href="../graph.assets/linkedlist_tree_graph.png" data-type="image" data-width="100%" data-height="auto" data-desc-position="bottom"><img alt="链表、树、图之间的关系" class="animation-figure" src="../graph.assets/linkedlist_tree_graph.png" /></a></p>
<p align="center"> 图 9-1 &nbsp; 链表、树、图之间的关系 </p>
<h2 id="911">9.1.1 &nbsp; 图常见类型与术语<a class="headerlink" href="#911" title="Permanent link">&para;</a></h2>
@@ -3424,7 +3424,7 @@ G &amp; = \{ V, E \} \newline
<li>在无向图中,边表示两顶点之间的“双向”连接关系,例如微信或 QQ 中的“好友关系”。</li>
<li>在有向图中,边具有方向性,即 <span class="arithmatex">\(A \rightarrow B\)</span><span class="arithmatex">\(A \leftarrow B\)</span> 两个方向的边是相互独立的,例如微博或抖音上的“关注”与“被关注”关系。</li>
</ul>
<p><a class="glightbox" href="../graph.assets/directed_graph.png" data-type="image" data-width="100%" data-height="auto" data-desc-position="bottom"><img alt="有向图与无向图" src="../graph.assets/directed_graph.png" /></a></p>
<p><a class="glightbox" href="../graph.assets/directed_graph.png" data-type="image" data-width="100%" data-height="auto" data-desc-position="bottom"><img alt="有向图与无向图" class="animation-figure" src="../graph.assets/directed_graph.png" /></a></p>
<p align="center"> 图 9-2 &nbsp; 有向图与无向图 </p>
<p>根据所有顶点是否连通,可分为图 9-3 所示的「连通图 connected graph」和「非连通图 disconnected graph」。</p>
@@ -3432,11 +3432,11 @@ G &amp; = \{ V, E \} \newline
<li>对于连通图,从某个顶点出发,可以到达其余任意顶点。</li>
<li>对于非连通图,从某个顶点出发,至少有一个顶点无法到达。</li>
</ul>
<p><a class="glightbox" href="../graph.assets/connected_graph.png" data-type="image" data-width="100%" data-height="auto" data-desc-position="bottom"><img alt="连通图与非连通图" src="../graph.assets/connected_graph.png" /></a></p>
<p><a class="glightbox" href="../graph.assets/connected_graph.png" data-type="image" data-width="100%" data-height="auto" data-desc-position="bottom"><img alt="连通图与非连通图" class="animation-figure" src="../graph.assets/connected_graph.png" /></a></p>
<p align="center"> 图 9-3 &nbsp; 连通图与非连通图 </p>
<p>我们还可以为边添加“权重”变量,从而得到图 9-4 所示的「有权图 weighted graph」。例如在王者荣耀等手游中系统会根据共同游戏时间来计算玩家之间的“亲密度”这种亲密度网络就可以用有权图来表示。</p>
<p><a class="glightbox" href="../graph.assets/weighted_graph.png" data-type="image" data-width="100%" data-height="auto" data-desc-position="bottom"><img alt="有权图与无权图" src="../graph.assets/weighted_graph.png" /></a></p>
<p><a class="glightbox" href="../graph.assets/weighted_graph.png" data-type="image" data-width="100%" data-height="auto" data-desc-position="bottom"><img alt="有权图与无权图" class="animation-figure" src="../graph.assets/weighted_graph.png" /></a></p>
<p align="center"> 图 9-4 &nbsp; 有权图与无权图 </p>
<p>图数据结构包含以下常用术语。</p>
@@ -3450,7 +3450,7 @@ G &amp; = \{ V, E \} \newline
<h3 id="1">1. &nbsp; 邻接矩阵<a class="headerlink" href="#1" title="Permanent link">&para;</a></h3>
<p>设图的顶点数量为 <span class="arithmatex">\(n\)</span> ,「邻接矩阵 adjacency matrix」使用一个 <span class="arithmatex">\(n \times n\)</span> 大小的矩阵来表示图,每一行(列)代表一个顶点,矩阵元素代表边,用 <span class="arithmatex">\(1\)</span><span class="arithmatex">\(0\)</span> 表示两个顶点之间是否存在边。</p>
<p>如图 9-5 所示,设邻接矩阵为 <span class="arithmatex">\(M\)</span>、顶点列表为 <span class="arithmatex">\(V\)</span> ,那么矩阵元素 <span class="arithmatex">\(M[i, j] = 1\)</span> 表示顶点 <span class="arithmatex">\(V[i]\)</span> 到顶点 <span class="arithmatex">\(V[j]\)</span> 之间存在边,反之 <span class="arithmatex">\(M[i, j] = 0\)</span> 表示两顶点之间无边。</p>
<p><a class="glightbox" href="../graph.assets/adjacency_matrix.png" data-type="image" data-width="100%" data-height="auto" data-desc-position="bottom"><img alt="图的邻接矩阵表示" src="../graph.assets/adjacency_matrix.png" /></a></p>
<p><a class="glightbox" href="../graph.assets/adjacency_matrix.png" data-type="image" data-width="100%" data-height="auto" data-desc-position="bottom"><img alt="图的邻接矩阵表示" class="animation-figure" src="../graph.assets/adjacency_matrix.png" /></a></p>
<p align="center"> 图 9-5 &nbsp; 图的邻接矩阵表示 </p>
<p>邻接矩阵具有以下特性。</p>
@@ -3462,7 +3462,7 @@ G &amp; = \{ V, E \} \newline
<p>使用邻接矩阵表示图时,我们可以直接访问矩阵元素以获取边,因此增删查操作的效率很高,时间复杂度均为 <span class="arithmatex">\(O(1)\)</span> 。然而,矩阵的空间复杂度为 <span class="arithmatex">\(O(n^2)\)</span> ,内存占用较多。</p>
<h3 id="2">2. &nbsp; 邻接表<a class="headerlink" href="#2" title="Permanent link">&para;</a></h3>
<p>「邻接表 adjacency list」使用 <span class="arithmatex">\(n\)</span> 个链表来表示图,链表节点表示顶点。第 <span class="arithmatex">\(i\)</span> 条链表对应顶点 <span class="arithmatex">\(i\)</span> ,其中存储了该顶点的所有邻接顶点(即与该顶点相连的顶点)。图 9-6 展示了一个使用邻接表存储的图的示例。</p>
<p><a class="glightbox" href="../graph.assets/adjacency_list.png" data-type="image" data-width="100%" data-height="auto" data-desc-position="bottom"><img alt="图的邻接表表示" src="../graph.assets/adjacency_list.png" /></a></p>
<p><a class="glightbox" href="../graph.assets/adjacency_list.png" data-type="image" data-width="100%" data-height="auto" data-desc-position="bottom"><img alt="图的邻接表表示" class="animation-figure" src="../graph.assets/adjacency_list.png" /></a></p>
<p align="center"> 图 9-6 &nbsp; 图的邻接表表示 </p>
<p>邻接表仅存储实际存在的边,而边的总数通常远小于 <span class="arithmatex">\(n^2\)</span> ,因此它更加节省空间。然而,在邻接表中需要通过遍历链表来查找边,因此其时间效率不如邻接矩阵。</p>