Skip to content

T
TRHX-github-io

weixin_48998305 / TRHX-github-io
与 Fork 源项目一致

Fork自 IT.BOB / TRHX-github-io

通知 1
Star 0
Fork 0
T
TRHX-github-io

前往新版Gitcode，体验更适合开发者的 AI 搜索 >>

提交 aed7b9cf 编写于 作者: T TRHX
浏览文件
操作

Site updated: 2020-08-07 09:16:17

上级 39168dc7
隐藏空白更改
内联 并排
Showing with 55 addition and 55 deletion
2018/08/10/A01-hello-world/index.html
浏览文件 @ aed7b9cf
......@@ -733,7 +733,7 @@
<h6 class="tags">
<a class="tag" href="/tags/Github-Pages/"><i class="fas fa-tags fa-fw" aria-hidden="true"></i>&nbsp;Github Pages</a> <a class="tag" href="/tags/Hexo/"><i class="fas fa-tags fa-fw" aria-hidden="true"></i>&nbsp;Hexo</a>
<a class="tag" href="/tags/Hexo/"><i class="fas fa-tags fa-fw" aria-hidden="true"></i>&nbsp;Hexo</a> <a class="tag" href="/tags/Github-Pages/"><i class="fas fa-tags fa-fw" aria-hidden="true"></i>&nbsp;Github Pages</a>
</h6>
</span>
......
2018/08/15/A02-hexo-blog/index.html
浏览文件 @ aed7b9cf
......@@ -747,7 +747,7 @@
<div class="new-meta-item meta-tags"><a class="tag" href="/tags/Github-Pages/" rel="nofollow"><i class="fas fa-tags" aria-hidden="true"></i>&nbsp;<p>Github Pages</p></a></div> <div class="new-meta-item meta-tags"><a class="tag" href="/tags/Hexo/" rel="nofollow"><i class="fas fa-tags" aria-hidden="true"></i>&nbsp;<p>Hexo</p></a></div>
<div class="new-meta-item meta-tags"><a class="tag" href="/tags/Hexo/" rel="nofollow"><i class="fas fa-tags" aria-hidden="true"></i>&nbsp;<p>Hexo</p></a></div> <div class="new-meta-item meta-tags"><a class="tag" href="/tags/Github-Pages/" rel="nofollow"><i class="fas fa-tags" aria-hidden="true"></i>&nbsp;<p>Github Pages</p></a></div>
......
2018/08/25/A03-markdown/index.html
浏览文件 @ aed7b9cf
......@@ -736,7 +736,7 @@
<h6 class="tags">
<a class="tag" href="/tags/Hexo/"><i class="fas fa-tags fa-fw" aria-hidden="true"></i>&nbsp;Hexo</a> <a class="tag" href="/tags/主题个性化/"><i class="fas fa-tags fa-fw" aria-hidden="true"></i>&nbsp;主题个性化</a> <a class="tag" href="/tags/Material-X/"><i class="fas fa-tags fa-fw" aria-hidden="true"></i>&nbsp;Material X</a> <a class="tag" href="/tags/spfk/"><i class="fas fa-tags fa-fw" aria-hidden="true"></i>&nbsp;spfk</a>
<a class="tag" href="/tags/主题个性化/"><i class="fas fa-tags fa-fw" aria-hidden="true"></i>&nbsp;主题个性化</a> <a class="tag" href="/tags/Hexo/"><i class="fas fa-tags fa-fw" aria-hidden="true"></i>&nbsp;Hexo</a> <a class="tag" href="/tags/Material-X/"><i class="fas fa-tags fa-fw" aria-hidden="true"></i>&nbsp;Material X</a> <a class="tag" href="/tags/spfk/"><i class="fas fa-tags fa-fw" aria-hidden="true"></i>&nbsp;spfk</a>
</h6>
</span>
......@@ -756,7 +756,7 @@
<h6 class="tags">
<a class="tag" href="/tags/Github-Pages/"><i class="fas fa-tags fa-fw" aria-hidden="true"></i>&nbsp;Github Pages</a> <a class="tag" href="/tags/Hexo/"><i class="fas fa-tags fa-fw" aria-hidden="true"></i>&nbsp;Hexo</a>
<a class="tag" href="/tags/Hexo/"><i class="fas fa-tags fa-fw" aria-hidden="true"></i>&nbsp;Hexo</a> <a class="tag" href="/tags/Github-Pages/"><i class="fas fa-tags fa-fw" aria-hidden="true"></i>&nbsp;Github Pages</a>
</h6>
</span>
......
2018/08/27/A04-Hexo-blog-topic-personalization/index.html
浏览文件 @ aed7b9cf
......@@ -916,7 +916,7 @@
<div class="new-meta-item meta-tags"><a class="tag" href="/tags/Hexo/" rel="nofollow"><i class="fas fa-tags" aria-hidden="true"></i>&nbsp;<p>Hexo</p></a></div> <div class="new-meta-item meta-tags"><a class="tag" href="/tags/主题个性化/" rel="nofollow"><i class="fas fa-tags" aria-hidden="true"></i>&nbsp;<p>主题个性化</p></a></div> <div class="new-meta-item meta-tags"><a class="tag" href="/tags/Material-X/" rel="nofollow"><i class="fas fa-tags" aria-hidden="true"></i>&nbsp;<p>Material X</p></a></div> <div class="new-meta-item meta-tags"><a class="tag" href="/tags/spfk/" rel="nofollow"><i class="fas fa-tags" aria-hidden="true"></i>&nbsp;<p>spfk</p></a></div>
<div class="new-meta-item meta-tags"><a class="tag" href="/tags/主题个性化/" rel="nofollow"><i class="fas fa-tags" aria-hidden="true"></i>&nbsp;<p>主题个性化</p></a></div> <div class="new-meta-item meta-tags"><a class="tag" href="/tags/Hexo/" rel="nofollow"><i class="fas fa-tags" aria-hidden="true"></i>&nbsp;<p>Hexo</p></a></div> <div class="new-meta-item meta-tags"><a class="tag" href="/tags/Material-X/" rel="nofollow"><i class="fas fa-tags" aria-hidden="true"></i>&nbsp;<p>Material X</p></a></div> <div class="new-meta-item meta-tags"><a class="tag" href="/tags/spfk/" rel="nofollow"><i class="fas fa-tags" aria-hidden="true"></i>&nbsp;<p>spfk</p></a></div>
......
2018/08/29/A05-markdown-editor/index.html
浏览文件 @ aed7b9cf
......@@ -817,7 +817,7 @@
<h6 class="tags">
<a class="tag" href="/tags/Hexo/"><i class="fas fa-tags fa-fw" aria-hidden="true"></i>&nbsp;Hexo</a> <a class="tag" href="/tags/主题个性化/"><i class="fas fa-tags fa-fw" aria-hidden="true"></i>&nbsp;主题个性化</a> <a class="tag" href="/tags/Material-X/"><i class="fas fa-tags fa-fw" aria-hidden="true"></i>&nbsp;Material X</a> <a class="tag" href="/tags/spfk/"><i class="fas fa-tags fa-fw" aria-hidden="true"></i>&nbsp;spfk</a>
<a class="tag" href="/tags/主题个性化/"><i class="fas fa-tags fa-fw" aria-hidden="true"></i>&nbsp;主题个性化</a> <a class="tag" href="/tags/Hexo/"><i class="fas fa-tags fa-fw" aria-hidden="true"></i>&nbsp;Hexo</a> <a class="tag" href="/tags/Material-X/"><i class="fas fa-tags fa-fw" aria-hidden="true"></i>&nbsp;Material X</a> <a class="tag" href="/tags/spfk/"><i class="fas fa-tags fa-fw" aria-hidden="true"></i>&nbsp;spfk</a>
</h6>
</span>
......
2020/06/17/A84-Pandas-06/index.html
浏览文件 @ aed7b9cf
......@@ -638,7 +638,7 @@
<p>对于多层分组，元组的第一个元素将会是由键值组成的元组，第二个元素为数据块：</p>
<figure class="highlight python"><table><tr><td class="gutter"><pre><span class="line">1</span><br><span class="line">2</span><br><span class="line">3</span><br><span class="line">4</span><br><span class="line">5</span><br><span class="line">6</span><br><span class="line">7</span><br><span class="line">8</span><br><span class="line">9</span><br><span class="line">10</span><br><span class="line">11</span><br><span class="line">12</span><br><span class="line">13</span><br><span class="line">14</span><br><span class="line">15</span><br><span class="line">16</span><br><span class="line">17</span><br><span class="line">18</span><br><span class="line">19</span><br><span class="line">20</span><br><span class="line">21</span><br><span class="line">22</span><br><span class="line">23</span><br><span class="line">24</span><br><span class="line">25</span><br><span class="line">26</span><br><span class="line">27</span><br><span class="line">28</span><br><span class="line">29</span><br><span class="line">30</span><br><span class="line">31</span><br><span class="line">32</span><br><span class="line">33</span><br><span class="line">34</span><br><span class="line">35</span><br><span class="line">36</span><br><span class="line">37</span><br><span class="line">38</span><br><span class="line">39</span><br><span class="line">40</span><br><span class="line">41</span><br><span class="line">42</span><br><span class="line">43</span><br></pre></td><td class="code"><pre><span class="line"><span class="meta">&gt;&gt;&gt; </span><span class="keyword">import</span> pandas <span class="keyword">as</span> pd</span><br><span class="line"><span class="meta">&gt;&gt;&gt; </span><span class="keyword">import</span> numpy <span class="keyword">as</span> np</span><br><span class="line"><span class="meta">&gt;&gt;&gt; </span>data = &#123;<span class="string">'key1'</span> : [<span class="string">'a'</span>, <span class="string">'b'</span>, <span class="string">'a'</span>, <span class="string">'b'</span>, <span class="string">'a'</span>, <span class="string">'b'</span>, <span class="string">'a'</span>, <span class="string">'a'</span>],</span><br><span class="line"> <span class="string">'key2'</span> : [<span class="string">'one'</span>, <span class="string">'one'</span>, <span class="string">'two'</span>, <span class="string">'three'</span>, <span class="string">'two'</span>, <span class="string">'two'</span>, <span class="string">'one'</span>, <span class="string">'three'</span>],</span><br><span class="line"> <span class="string">'data1'</span>: np.random.randn(<span class="number">8</span>),</span><br><span class="line"> <span class="string">'data2'</span>: np.random.randn(<span class="number">8</span>)&#125;</span><br><span class="line"><span class="meta">&gt;&gt;&gt; </span>obj = pd.DataFrame(data)</span><br><span class="line"><span class="meta">&gt;&gt;&gt; </span>obj</span><br><span class="line"> key1 key2 data1 data2</span><br><span class="line"><span class="number">0</span> a one <span class="number">-1.088762</span> <span class="number">0.668504</span></span><br><span class="line"><span class="number">1</span> b one <span class="number">0.275500</span> <span class="number">0.787844</span></span><br><span class="line"><span class="number">2</span> a two <span class="number">-0.108417</span> <span class="number">-0.491296</span></span><br><span class="line"><span class="number">3</span> b three <span class="number">0.019524</span> <span class="number">-0.363390</span></span><br><span class="line"><span class="number">4</span> a two <span class="number">0.453612</span> <span class="number">0.796999</span></span><br><span class="line"><span class="number">5</span> b two <span class="number">1.982858</span> <span class="number">1.501877</span></span><br><span class="line"><span class="number">6</span> a one <span class="number">1.101132</span> <span class="number">-1.928362</span></span><br><span class="line"><span class="number">7</span> a three <span class="number">0.524775</span> <span class="number">-1.205842</span></span><br><span class="line"><span class="meta">&gt;&gt;&gt; </span></span><br><span class="line"><span class="meta">&gt;&gt;&gt; </span><span class="keyword">for</span> group_name, group_data <span class="keyword">in</span> obj.groupby([<span class="string">'key1'</span>, <span class="string">'key2'</span>]):</span><br><span class="line"> print(group_name)</span><br><span class="line"> print(group_data)</span><br><span class="line"></span><br><span class="line"> </span><br><span class="line">(<span class="string">'a'</span>, <span class="string">'one'</span>)</span><br><span class="line"> key1 key2 data1 data2</span><br><span class="line"><span class="number">0</span> a one <span class="number">-1.088762</span> <span class="number">0.668504</span></span><br><span class="line"><span class="number">6</span> a one <span class="number">1.101132</span> <span class="number">-1.928362</span></span><br><span class="line">(<span class="string">'a'</span>, <span class="string">'three'</span>)</span><br><span class="line"> key1 key2 data1 data2</span><br><span class="line"><span class="number">7</span> a three <span class="number">0.524775</span> <span class="number">-1.205842</span></span><br><span class="line">(<span class="string">'a'</span>, <span class="string">'two'</span>)</span><br><span class="line"> key1 key2 data1 data2</span><br><span class="line"><span class="number">2</span> a two <span class="number">-0.108417</span> <span class="number">-0.491296</span></span><br><span class="line"><span class="number">4</span> a two <span class="number">0.453612</span> <span class="number">0.796999</span></span><br><span class="line">(<span class="string">'b'</span>, <span class="string">'one'</span>)</span><br><span class="line"> key1 key2 data1 data2</span><br><span class="line"><span class="number">1</span> b one <span class="number">0.2755</span> <span class="number">0.787844</span></span><br><span class="line">(<span class="string">'b'</span>, <span class="string">'three'</span>)</span><br><span class="line"> key1 key2 data1 data2</span><br><span class="line"><span class="number">3</span> b three <span class="number">0.019524</span> <span class="number">-0.36339</span></span><br><span class="line">(<span class="string">'b'</span>, <span class="string">'two'</span>)</span><br><span class="line"> key1 key2 data1 data2</span><br><span class="line"><span class="number">5</span> b two <span class="number">1.982858</span> <span class="number">1.501877</span></span><br></pre></td></tr></table></figure>
<h3 id="【03x05】对象转换"><a href="#【03x05】对象转换" class="headerlink" title="【03x05】对象转换"></a><font color="#4876FF">【03x05】对象转换</font></h3><p>GroupBy 对象支持转换成列表或字典：</p>
<figure class="highlight python"><table><tr><td class="gutter"><pre><span class="line">1</span><br><span class="line">2</span><br><span class="line">3</span><br><span class="line">4</span><br><span class="line">5</span><br><span class="line">6</span><br><span class="line">7</span><br><span class="line">8</span><br><span class="line">9</span><br><span class="line">10</span><br><span class="line">11</span><br><span class="line">12</span><br><span class="line">13</span><br><span class="line">14</span><br><span class="line">15</span><br><span class="line">16</span><br><span class="line">17</span><br><span class="line">18</span><br><span class="line">19</span><br><span class="line">20</span><br><span class="line">21</span><br><span class="line">22</span><br><span class="line">23</span><br><span class="line">24</span><br><span class="line">25</span><br><span class="line">26</span><br><span class="line">27</span><br><span class="line">28</span><br><span class="line">29</span><br><span class="line">30</span><br><span class="line">31</span><br><span class="line">32</span><br><span class="line">33</span><br><span class="line">34</span><br><span class="line">35</span><br><span class="line">36</span><br><span class="line">37</span><br><span class="line">38</span><br><span class="line">39</span><br><span class="line">40</span><br><span class="line">41</span><br><span class="line">42</span><br></pre></td><td class="code"><pre><span class="line"><span class="meta">&gt;&gt;&gt; </span><span class="keyword">import</span> pandas <span class="keyword">as</span> pd</span><br><span class="line"><span class="meta">&gt;&gt;&gt; </span><span class="keyword">import</span> numpy <span class="keyword">as</span> np</span><br><span class="line"><span class="meta">&gt;&gt;&gt; </span>data = &#123;<span class="string">'key1'</span> : [<span class="string">'a'</span>, <span class="string">'b'</span>, <span class="string">'a'</span>, <span class="string">'b'</span>, <span class="string">'a'</span>, <span class="string">'b'</span>, <span class="string">'a'</span>, <span class="string">'a'</span>],</span><br><span class="line"> <span class="string">'key2'</span> : [<span class="string">'one'</span>, <span class="string">'one'</span>, <span class="string">'two'</span>, <span class="string">'three'</span>, <span class="string">'two'</span>, <span class="string">'two'</span>, <span class="string">'one'</span>, <span class="string">'three'</span>],</span><br><span class="line"> <span class="string">'data1'</span>: np.random.randn(<span class="number">8</span>),</span><br><span class="line"> <span class="string">'data2'</span>: np.random.randn(<span class="number">8</span>)&#125;</span><br><span class="line"><span class="meta">&gt;&gt;&gt; </span>obj = pd.DataFrame(data)</span><br><span class="line"><span class="meta">&gt;&gt;&gt; </span>obj</span><br><span class="line"> key1 key2 data1 data2</span><br><span class="line"><span class="number">0</span> a one <span class="number">-0.607009</span> <span class="number">1.948301</span></span><br><span class="line"><span class="number">1</span> b one <span class="number">0.150818</span> <span class="number">-0.025095</span></span><br><span class="line"><span class="number">2</span> a two <span class="number">-2.086024</span> <span class="number">0.358164</span></span><br><span class="line"><span class="number">3</span> b three <span class="number">0.446061</span> <span class="number">1.708797</span></span><br><span class="line"><span class="number">4</span> a two <span class="number">0.745457</span> <span class="number">-0.980948</span></span><br><span class="line"><span class="number">5</span> b two <span class="number">0.981877</span> <span class="number">2.159327</span></span><br><span class="line"><span class="number">6</span> a one <span class="number">0.804480</span> <span class="number">-0.499661</span></span><br><span class="line"><span class="number">7</span> a three <span class="number">0.112884</span> <span class="number">0.004367</span></span><br><span class="line"><span class="meta">&gt;&gt;&gt; </span></span><br><span class="line"><span class="meta">&gt;&gt;&gt; </span>grouped = obj.groupby(<span class="string">'key1'</span>)</span><br><span class="line"><span class="meta">&gt;&gt;&gt; </span>list(grouped1)</span><br><span class="line">[(<span class="string">'a'</span>, key1 key2 data1 data2</span><br><span class="line"><span class="number">0</span> a one <span class="number">-0.607009</span> <span class="number">1.948301</span></span><br><span class="line"><span class="number">2</span> a two <span class="number">-2.086024</span> <span class="number">0.358164</span></span><br><span class="line"><span class="number">4</span> a two <span class="number">0.745457</span> <span class="number">-0.980948</span></span><br><span class="line"><span class="number">6</span> a one <span class="number">0.804480</span> <span class="number">-0.499661</span></span><br><span class="line"><span class="number">7</span> a three <span class="number">0.112884</span> <span class="number">0.004367</span>),</span><br><span class="line">(<span class="string">'b'</span>, key1 key2 data1 data2</span><br><span class="line"><span class="number">1</span> b one <span class="number">0.150818</span> <span class="number">-0.025095</span></span><br><span class="line"><span class="number">3</span> b three <span class="number">0.446061</span> <span class="number">1.708797</span></span><br><span class="line"><span class="number">5</span> b two <span class="number">0.981877</span> <span class="number">2.159327</span>)]</span><br><span class="line"><span class="meta">&gt;&gt;&gt; </span></span><br><span class="line"><span class="meta">&gt;&gt;&gt; </span>dict(list(grouped1))</span><br><span class="line">&#123;<span class="string">'a'</span>: key1 key2 data1 data2</span><br><span class="line"><span class="number">0</span> a one <span class="number">-0.607009</span> <span class="number">1.948301</span></span><br><span class="line"><span class="number">2</span> a two <span class="number">-2.086024</span> <span class="number">0.358164</span></span><br><span class="line"><span class="number">4</span> a two <span class="number">0.745457</span> <span class="number">-0.980948</span></span><br><span class="line"><span class="number">6</span> a one <span class="number">0.804480</span> <span class="number">-0.499661</span></span><br><span class="line"><span class="number">7</span> a three <span class="number">0.112884</span> <span class="number">0.004367</span>,</span><br><span class="line"><span class="string">'b'</span>: key1 key2 data1 data2</span><br><span class="line"><span class="number">1</span> b one <span class="number">0.150818</span> <span class="number">-0.025095</span></span><br><span class="line"><span class="number">3</span> b three <span class="number">0.446061</span> <span class="number">1.708797</span></span><br><span class="line"><span class="number">5</span> b two <span class="number">0.981877</span> <span class="number">2.159327</span>&#125;</span><br></pre></td></tr></table></figure>
<figure class="highlight python"><table><tr><td class="gutter"><pre><span class="line">1</span><br><span class="line">2</span><br><span class="line">3</span><br><span class="line">4</span><br><span class="line">5</span><br><span class="line">6</span><br><span class="line">7</span><br><span class="line">8</span><br><span class="line">9</span><br><span class="line">10</span><br><span class="line">11</span><br><span class="line">12</span><br><span class="line">13</span><br><span class="line">14</span><br><span class="line">15</span><br><span class="line">16</span><br><span class="line">17</span><br><span class="line">18</span><br><span class="line">19</span><br><span class="line">20</span><br><span class="line">21</span><br><span class="line">22</span><br><span class="line">23</span><br><span class="line">24</span><br><span class="line">25</span><br><span class="line">26</span><br><span class="line">27</span><br><span class="line">28</span><br><span class="line">29</span><br><span class="line">30</span><br><span class="line">31</span><br><span class="line">32</span><br><span class="line">33</span><br><span class="line">34</span><br><span class="line">35</span><br><span class="line">36</span><br><span class="line">37</span><br><span class="line">38</span><br><span class="line">39</span><br><span class="line">40</span><br><span class="line">41</span><br><span class="line">42</span><br></pre></td><td class="code"><pre><span class="line"><span class="meta">&gt;&gt;&gt; </span><span class="keyword">import</span> pandas <span class="keyword">as</span> pd</span><br><span class="line"><span class="meta">&gt;&gt;&gt; </span><span class="keyword">import</span> numpy <span class="keyword">as</span> np</span><br><span class="line"><span class="meta">&gt;&gt;&gt; </span>data = &#123;<span class="string">'key1'</span> : [<span class="string">'a'</span>, <span class="string">'b'</span>, <span class="string">'a'</span>, <span class="string">'b'</span>, <span class="string">'a'</span>, <span class="string">'b'</span>, <span class="string">'a'</span>, <span class="string">'a'</span>],</span><br><span class="line"> <span class="string">'key2'</span> : [<span class="string">'one'</span>, <span class="string">'one'</span>, <span class="string">'two'</span>, <span class="string">'three'</span>, <span class="string">'two'</span>, <span class="string">'two'</span>, <span class="string">'one'</span>, <span class="string">'three'</span>],</span><br><span class="line"> <span class="string">'data1'</span>: np.random.randn(<span class="number">8</span>),</span><br><span class="line"> <span class="string">'data2'</span>: np.random.randn(<span class="number">8</span>)&#125;</span><br><span class="line"><span class="meta">&gt;&gt;&gt; </span>obj = pd.DataFrame(data)</span><br><span class="line"><span class="meta">&gt;&gt;&gt; </span>obj</span><br><span class="line"> key1 key2 data1 data2</span><br><span class="line"><span class="number">0</span> a one <span class="number">-0.607009</span> <span class="number">1.948301</span></span><br><span class="line"><span class="number">1</span> b one <span class="number">0.150818</span> <span class="number">-0.025095</span></span><br><span class="line"><span class="number">2</span> a two <span class="number">-2.086024</span> <span class="number">0.358164</span></span><br><span class="line"><span class="number">3</span> b three <span class="number">0.446061</span> <span class="number">1.708797</span></span><br><span class="line"><span class="number">4</span> a two <span class="number">0.745457</span> <span class="number">-0.980948</span></span><br><span class="line"><span class="number">5</span> b two <span class="number">0.981877</span> <span class="number">2.159327</span></span><br><span class="line"><span class="number">6</span> a one <span class="number">0.804480</span> <span class="number">-0.499661</span></span><br><span class="line"><span class="number">7</span> a three <span class="number">0.112884</span> <span class="number">0.004367</span></span><br><span class="line"><span class="meta">&gt;&gt;&gt; </span></span><br><span class="line"><span class="meta">&gt;&gt;&gt; </span>grouped = obj.groupby(<span class="string">'key1'</span>)</span><br><span class="line"><span class="meta">&gt;&gt;&gt; </span>list(grouped)</span><br><span class="line">[(<span class="string">'a'</span>, key1 key2 data1 data2</span><br><span class="line"><span class="number">0</span> a one <span class="number">-0.607009</span> <span class="number">1.948301</span></span><br><span class="line"><span class="number">2</span> a two <span class="number">-2.086024</span> <span class="number">0.358164</span></span><br><span class="line"><span class="number">4</span> a two <span class="number">0.745457</span> <span class="number">-0.980948</span></span><br><span class="line"><span class="number">6</span> a one <span class="number">0.804480</span> <span class="number">-0.499661</span></span><br><span class="line"><span class="number">7</span> a three <span class="number">0.112884</span> <span class="number">0.004367</span>),</span><br><span class="line">(<span class="string">'b'</span>, key1 key2 data1 data2</span><br><span class="line"><span class="number">1</span> b one <span class="number">0.150818</span> <span class="number">-0.025095</span></span><br><span class="line"><span class="number">3</span> b three <span class="number">0.446061</span> <span class="number">1.708797</span></span><br><span class="line"><span class="number">5</span> b two <span class="number">0.981877</span> <span class="number">2.159327</span>)]</span><br><span class="line">&gt;&gt;&gt;</span><br><span class="line"><span class="meta">&gt;&gt;&gt; </span>dict(list(grouped))</span><br><span class="line">&#123;<span class="string">'a'</span>: key1 key2 data1 data2</span><br><span class="line"><span class="number">0</span> a one <span class="number">-0.607009</span> <span class="number">1.948301</span></span><br><span class="line"><span class="number">2</span> a two <span class="number">-2.086024</span> <span class="number">0.358164</span></span><br><span class="line"><span class="number">4</span> a two <span class="number">0.745457</span> <span class="number">-0.980948</span></span><br><span class="line"><span class="number">6</span> a one <span class="number">0.804480</span> <span class="number">-0.499661</span></span><br><span class="line"><span class="number">7</span> a three <span class="number">0.112884</span> <span class="number">0.004367</span>,</span><br><span class="line"><span class="string">'b'</span>: key1 key2 data1 data2</span><br><span class="line"><span class="number">1</span> b one <span class="number">0.150818</span> <span class="number">-0.025095</span></span><br><span class="line"><span class="number">3</span> b three <span class="number">0.446061</span> <span class="number">1.708797</span></span><br><span class="line"><span class="number">5</span> b two <span class="number">0.981877</span> <span class="number">2.159327</span>&#125;</span><br></pre></td></tr></table></figure>
<h2 id="【04x00】GroupBy-Apply-数据应用"><a href="#【04x00】GroupBy-Apply-数据应用" class="headerlink" title="【04x00】GroupBy Apply 数据应用"></a><font color="#FF0000">【04x00】GroupBy Apply 数据应用</font></h2><p>聚合指的是任何能够从数组产生标量值的数据转换过程，常用于对分组后的数据进行计算</p>
<h3 id="【04x01】聚合函数"><a href="#【04x01】聚合函数" class="headerlink" title="【04x01】聚合函数"></a><font color="#4876FF">【04x01】聚合函数</font></h3><p>之前的例子已经用过一些内置的聚合函数，比如 mean、count、min 以及 sum 等。常见的聚合运算如下表所示：</p>
<p>官方文档：<a href="https://pandas.pydata.org/docs/reference/groupby.html" target="_blank" rel="noopener">https://pandas.pydata.org/docs/reference/groupby.html</a></p>
......@@ -807,10 +807,10 @@
<div class='new-meta-box'>
<div class="new-meta-item date" itemprop="dateUpdated" datetime="2020-08-06T11:31:25+08:00">
<div class="new-meta-item date" itemprop="dateUpdated" datetime="2020-08-07T09:13:45+08:00">
<a class='notlink'>
<i class="fas fa-clock" aria-hidden="true"></i>
<p>最后更新于 2020年8月6日</p>
<p>最后更新于 2020年8月7日</p>
</a>
</div>
......
archives/2018/08/index.html
浏览文件 @ aed7b9cf
......@@ -682,10 +682,10 @@
<div class="full-width auto-padding tags">
<a href="/tags/Hexo/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;Hexo</a>
<a href="/tags/主题个性化/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;主题个性化</a>
<a href="/tags/Hexo/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;Hexo</a>
<a href="/tags/Material-X/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;Material X</a>
<a href="/tags/spfk/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;spfk</a>
......@@ -961,10 +961,10 @@
<div class="full-width auto-padding tags">
<a href="/tags/Github-Pages/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;Github Pages</a>
<a href="/tags/Hexo/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;Hexo</a>
<a href="/tags/Github-Pages/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;Github Pages</a>
</div>
</section>
......
archives/2018/page/2/index.html
浏览文件 @ aed7b9cf
......@@ -668,10 +668,10 @@
<div class="full-width auto-padding tags">
<a href="/tags/Hexo/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;Hexo</a>
<a href="/tags/主题个性化/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;主题个性化</a>
<a href="/tags/Hexo/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;Hexo</a>
<a href="/tags/Material-X/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;Material X</a>
<a href="/tags/spfk/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;spfk</a>
......@@ -947,10 +947,10 @@
<div class="full-width auto-padding tags">
<a href="/tags/Github-Pages/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;Github Pages</a>
<a href="/tags/Hexo/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;Hexo</a>
<a href="/tags/Github-Pages/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;Github Pages</a>
</div>
</section>
......
atom.xml
浏览文件 @ aed7b9cf
因为 它太大了无法显示 source diff 。你可以改为 查看blob。
baidusitemap.xml
浏览文件 @ aed7b9cf
<?xml version="1.0" encoding="UTF-8"?>
<urlset xmlns="http://www.sitemaps.org/schemas/sitemap/0.9">
<url>
<loc>https://www.itrhx.com/2020/06/17/A84-Pandas-06/</loc>
<lastmod>2020-08-07</lastmod>
</url> <url>
<loc>https://www.itrhx.com/2020/07/13/A90-pyspider-51job/</loc>
<lastmod>2020-08-06</lastmod>
</url> <url>
......@@ -9,9 +12,6 @@
</url> <url>
<loc>https://www.itrhx.com/2020/06/26/A88-Pandas-10/</loc>
<lastmod>2020-08-06</lastmod>
</url> <url>
<loc>https://www.itrhx.com/2020/06/17/A84-Pandas-06/</loc>
<lastmod>2020-08-06</lastmod>
</url> <url>
<loc>https://www.itrhx.com/2020/06/11/A79-Pandas-01/</loc>
<lastmod>2020-08-06</lastmod>
......
categories/Hexo/index.html
浏览文件 @ aed7b9cf
......@@ -1473,10 +1473,10 @@
<div class="full-width auto-padding tags">
<a href="/tags/Hexo/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;Hexo</a>
<a href="/tags/主题个性化/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;主题个性化</a>
<a href="/tags/Hexo/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;Hexo</a>
<a href="/tags/Material-X/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;Material X</a>
<a href="/tags/spfk/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;spfk</a>
......@@ -1623,10 +1623,10 @@
<div class="full-width auto-padding tags">
<a href="/tags/Github-Pages/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;Github Pages</a>
<a href="/tags/Hexo/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;Hexo</a>
<a href="/tags/Github-Pages/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;Github Pages</a>
</div>
</section>
......
content.json
浏览文件 @ aed7b9cf
因为 它太大了无法显示 source diff 。你可以改为 查看blob。
page/9/index.html
浏览文件 @ aed7b9cf
......@@ -1229,10 +1229,10 @@
<div class="full-width auto-padding tags">
<a href="/tags/Hexo/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;Hexo</a>
<a href="/tags/主题个性化/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;主题个性化</a>
<a href="/tags/Hexo/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;Hexo</a>
<a href="/tags/Material-X/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;Material X</a>
<a href="/tags/spfk/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;spfk</a>
......@@ -1508,10 +1508,10 @@
<div class="full-width auto-padding tags">
<a href="/tags/Github-Pages/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;Github Pages</a>
<a href="/tags/Hexo/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;Hexo</a>
<a href="/tags/Github-Pages/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;Github Pages</a>
</div>
</section>
......
search.xml
浏览文件 @ aed7b9cf
......@@ -84,7 +84,7 @@
<entry>
<title><![CDATA[Python 数据分析三剑客之 Pandas（六）：GroupBy 数据分裂/应用/合并]]></title>
<url>%2F2020%2F06%2F17%2FA84-Pandas-06%2F</url>
<content type="text"><![CDATA[Pandas 系列文章： Python 数据分析三剑客之 Pandas（一）：认识 Pandas 及其 Series、DataFrame 对象 Python 数据分析三剑客之 Pandas（二）：Index 索引对象以及各种索引操作 Python 数据分析三剑客之 Pandas（三）：算术运算与缺失值的处理 Python 数据分析三剑客之 Pandas（四）：函数应用、映射、排序和层级索引 Python 数据分析三剑客之 Pandas（五）：统计计算与统计描述 Python 数据分析三剑客之 Pandas（六）：GroupBy 数据分裂、应用与合并 Python 数据分析三剑客之 Pandas（七）：合并数据集 Python 数据分析三剑客之 Pandas（八）：数据重塑、重复数据处理与数据替换 Python 数据分析三剑客之 Pandas（九）：时间序列 Python 数据分析三剑客之 Pandas（十）：数据读写 专栏： 【NumPy 专栏】【Pandas 专栏】【Matplotlib 专栏】 推荐学习资料与网站： 【NumPy 中文网】【Pandas 中文网】【Matplotlib 中文网】【NumPy、Matplotlib、Pandas 速查表】 12345这里是一段防爬虫文本，请读者忽略。本文原创首发于 CSDN，作者 TRHX。博客首页：https://itrhx.blog.csdn.net/本文链接：https://itrhx.blog.csdn.net/article/details/106804881未经授权，禁止转载！恶意转载，后果自负！尊重原创，远离剽窃！ 【01x00】GroupBy 机制对数据集进行分组并对各组应用一个函数（无论是聚合还是转换），通常是数据分析工作中的重要环节。在将数据集加载、融合、准备好之后，通常就是计算分组统计或生成透视表。Pandas 提供了一个灵活高效的 GroupBy 功能，虽然“分组”（group by）这个名字是借用 SQL 数据库语言的命令，但其理念引用发明 R 语言 frame 的 Hadley Wickham 的观点可能更合适：分裂（Split）、应用（Apply）和组合（Combine）。 分组运算过程：Split —&gt; Apply —&gt; Combine 分裂（Split）：根据某些标准将数据分组； 应用（Apply）：对每个组独立应用一个函数； 合并（Combine）：把每个分组的计算结果合并起来。 官方介绍：https://pandas.pydata.org/docs/user_guide/groupby.html 【02x00】GroupBy 对象常见的 GroupBy 对象：Series.groupby、DataFrame.groupby，基本语法如下： 123456789Series.groupby(self, by=None, axis=0, level=None, as_index: bool = True, sort: bool = True, group_keys: bool = True, squeeze: bool = False, observed: bool = False) → ’groupby_generic.SeriesGroupBy’ 123456789DataFrame.groupby(self, by=None, axis=0, level=None, as_index: bool = True, sort: bool = True, group_keys: bool = True, squeeze: bool = False, observed: bool = False) → ’groupby_generic.DataFrameGroupBy’ 官方文档： https://pandas.pydata.org/docs/reference/api/pandas.Series.groupby.html https://pandas.pydata.org/docs/reference/api/pandas.DataFrame.groupby.html 常用参数解释如下： 参数 描述 by 映射、函数、标签或标签列表，用于确定分组依据的分组。如果 by 是函数，则会在对象索引的每个值上调用它。 如果传递了 dict 或 Series，则将使用 Series 或 dict 的值来确定组（将 Series 的值首先对齐；请参见.align() 方法）。 如果传递了 ndarray，则按原样使用这些值来确定组。标签或标签列表可以按自身中的列传递给分组。 注意，元组被解释为（单个）键 axis 沿指定轴拆分，默认 0，0 or ‘index’，1 or ‘columns’，只有在 DataFrame 中才有 1 or &#39;columns’ level 如果轴是 MultiIndex（层次结构），则按特定层级进行分组，默认 None as_index bool 类型，默认 True，对于聚合输出，返回以组标签为索引的对象。仅与 DataFrame 输入相关。as_index=False 实际上是“SQL样式”分组输出 sort bool 类型，默认 True，对组键排序。关闭此选项可获得更好的性能。注：这不影响每组的观察顺序。Groupby 保留每个组中行的顺序 group_keys bool 类型，默认 True，调用 apply 方法时，是否将组键（keys）添加到索引（ index）以标识块 squeeze bool 类型，默认 False，如果可能，减少返回类型的维度，否则返回一致的类型 groupby() 进行分组，GroupBy 对象没有进行实际运算，只是包含分组的中间数据，示例如下： 123456789101112131415161718192021222324&gt;&gt;&gt; import pandas as pd&gt;&gt;&gt; import numpy as np&gt;&gt;&gt; data = &#123;'key1' : ['a', 'b', 'a', 'b', 'a', 'b', 'a', 'a'], 'key2' : ['one', 'one', 'two', 'three', 'two', 'two', 'one', 'three'], 'data1': np.random.randn(8), 'data2': np.random.randn(8)&#125;&gt;&gt;&gt; &gt;&gt;&gt; obj = pd.DataFrame(data)&gt;&gt;&gt; obj key1 key2 data1 data20 a one -0.804160 -0.8689051 b one -0.086990 0.3257412 a two 0.757992 0.5411013 b three -0.281435 0.0978414 a two 0.817757 -0.6436995 b two -0.462760 -0.3211966 a one -0.403699 0.6021387 a three 0.883940 -0.850526&gt;&gt;&gt; &gt;&gt;&gt; obj.groupby('key1')&lt;pandas.core.groupby.generic.DataFrameGroupBy object at 0x03CDB7C0&gt;&gt;&gt;&gt; &gt;&gt;&gt; obj['data1'].groupby(obj['key1'])&lt;pandas.core.groupby.generic.SeriesGroupBy object at 0x03CDB748&gt; 【03x00】GroupBy Split 数据分裂【03x01】分组运算前面通过 groupby() 方法获得了一个 GroupBy 对象，它实际上还没有进行任何计算，只是含有一些有关分组键 obj[&#39;key1&#39;] 的中间数据而已。换句话说，该对象已经有了接下来对各分组执行运算所需的一切信息。例如，我们可以调用 GroupBy 的 mean() 方法来计算分组平均值，size() 方法返回每个分组的元素个数： 123456789101112131415161718192021222324252627282930313233343536373839404142434445&gt;&gt;&gt; import pandas as pd&gt;&gt;&gt; import numpy as np&gt;&gt;&gt; data = &#123;'key1' : ['a', 'b', 'a', 'b', 'a', 'b', 'a', 'a'], 'key2' : ['one', 'one', 'two', 'three', 'two', 'two', 'one', 'three'], 'data1': np.random.randn(8), 'data2': np.random.randn(8)&#125;&gt;&gt;&gt; &gt;&gt;&gt; obj = pd.DataFrame(data)&gt;&gt;&gt; obj key1 key2 data1 data20 a one -0.544099 -0.6140791 b one 2.193712 0.1010052 a two -0.004683 0.8827703 b three 0.312858 1.7321054 a two 0.011089 0.0895875 b two 0.292165 1.3276386 a one -1.433291 -0.2389717 a three -0.004724 -2.117326&gt;&gt;&gt; &gt;&gt;&gt; grouped1 = obj.groupby('key1')&gt;&gt;&gt; grouped2 = obj['data1'].groupby(obj['key1'])&gt;&gt;&gt; &gt;&gt;&gt; grouped1.mean() data1 data2key1 a -0.395142 -0.399604b 0.932912 1.053583&gt;&gt;&gt; &gt;&gt;&gt; grouped2.mean()key1a -0.395142b 0.932912Name: data1, dtype: float64&gt;&gt;&gt;&gt;&gt;&gt; grouped1.size()key1a 5b 3dtype: int64&gt;&gt;&gt; &gt;&gt;&gt; grouped2.size()key1a 5b 3Name: data1, dtype: int64 【03x02】按类型按列分组groupby() 方法 axis 参数默认是 0，通过设置也可以在其他任何轴上进行分组，也支持按照类型（dtype）进行分组： 12345678910111213141516171819202122232425262728293031323334353637383940&gt;&gt;&gt; import pandas as pd&gt;&gt;&gt; import numpy as np&gt;&gt;&gt; data = &#123;'key1' : ['a', 'b', 'a', 'b', 'a', 'b', 'a', 'a'], 'key2' : ['one', 'one', 'two', 'three', 'two', 'two', 'one', 'three'], 'data1': np.random.randn(8), 'data2': np.random.randn(8)&#125;&gt;&gt;&gt; obj = pd.DataFrame(data)&gt;&gt;&gt; obj key1 key2 data1 data20 a one -0.607009 1.9483011 b one 0.150818 -0.0250952 a two -2.086024 0.3581643 b three 0.446061 1.7087974 a two 0.745457 -0.9809485 b two 0.981877 2.1593276 a one 0.804480 -0.4996617 a three 0.112884 0.004367&gt;&gt;&gt; &gt;&gt;&gt; obj.dtypeskey1 objectkey2 objectdata1 float64data2 float64dtype: object&gt;&gt;&gt; &gt;&gt;&gt; obj.groupby(obj.dtypes, axis=1).size()float64 2object 2dtype: int64&gt;&gt;&gt; &gt;&gt;&gt; obj.groupby(obj.dtypes, axis=1).sum() float64 object0 1.341291 aone1 0.125723 bone2 -1.727860 atwo3 2.154858 bthree4 -0.235491 atwo5 3.141203 btwo6 0.304819 aone7 0.117251 athree 【03x03】自定义分组groupby() 方法中可以一次传入多个数组的列表，也可以自定义一组分组键。也可以通过一个字典、一个函数，或者按照索引层级进行分组。 传入多个数组的列表： 12345678910111213141516171819202122232425262728293031323334&gt;&gt;&gt; import pandas as pd&gt;&gt;&gt; import numpy as np&gt;&gt;&gt; data = &#123;'key1' : ['a', 'b', 'a', 'b', 'a', 'b', 'a', 'a'], 'key2' : ['one', 'one', 'two', 'three', 'two', 'two', 'one', 'three'], 'data1': np.random.randn(8), 'data2': np.random.randn(8)&#125;&gt;&gt;&gt; obj = pd.DataFrame(data)&gt;&gt;&gt; obj key1 key2 data1 data20 a one -0.841652 0.6880551 b one 0.510042 -0.5611712 a two -0.418862 -0.1459833 b three -1.104698 0.5631584 a two 0.329527 -0.8931085 b two 0.753653 -0.3425206 a one -0.882527 -1.1213297 a three 1.726794 0.160244&gt;&gt;&gt; &gt;&gt;&gt; means = obj['data1'].groupby([obj['key1'], obj['key2']]).mean()&gt;&gt;&gt; meanskey1 key2 a one -0.862090 three 1.726794 two -0.044667b one 0.510042 three -1.104698 two 0.753653Name: data1, dtype: float64&gt;&gt;&gt; &gt;&gt;&gt; means.unstack()key2 one three twokey1 a -0.862090 1.726794 -0.044667b 0.510042 -1.104698 0.753653 自定义分组键： 1234567891011121314151617181920212223&gt;&gt;&gt; import pandas as pd&gt;&gt;&gt; import numpy as np&gt;&gt;&gt; obj = pd.DataFrame(&#123;'key1' : ['a', 'a', 'b', 'b', 'a'], 'key2' : ['one', 'two', 'one', 'two', 'one'], 'data1' : np.random.randn(5), 'data2' : np.random.randn(5)&#125;)&gt;&gt;&gt; obj key1 key2 data1 data20 a one -0.024003 0.3504801 a two -0.767534 -0.1004262 b one -0.594983 -1.9455803 b two -0.374482 0.8175924 a one 0.755452 -0.137759&gt;&gt;&gt; &gt;&gt;&gt; states = np.array(['Wuhan', 'Beijing', 'Beijing', 'Wuhan', 'Wuhan'])&gt;&gt;&gt; years = np.array([2005, 2005, 2006, 2005, 2006])&gt;&gt;&gt; &gt;&gt;&gt; obj['data1'].groupby([states, years]).mean()Beijing 2005 -0.767534 2006 -0.594983Wuhan 2005 -0.199242 2006 0.755452Name: data1, dtype: float64 【03x03x01】字典分组通过字典进行分组： 1234567891011121314151617181920212223242526272829303132333435&gt;&gt;&gt; import pandas as pd&gt;&gt;&gt; import numpy as np&gt;&gt;&gt; obj = pd.DataFrame(np.random.randint(1, 10, (5,5)), columns=['a', 'b', 'c', 'd', 'e'], index=['A', 'B', 'C', 'D', 'E'])&gt;&gt;&gt; obj a b c d eA 1 4 7 1 9B 8 2 4 7 8C 9 8 2 5 1D 2 4 2 8 3E 7 5 7 2 3&gt;&gt;&gt; &gt;&gt;&gt; obj_dict = &#123;'a':'Python', 'b':'Python', 'c':'Java', 'd':'C++', 'e':'Java'&#125;&gt;&gt;&gt; obj.groupby(obj_dict, axis=1).size()C++ 1Java 2Python 2dtype: int64&gt;&gt;&gt; &gt;&gt;&gt; obj.groupby(obj_dict, axis=1).count() C++ Java PythonA 1 2 2B 1 2 2C 1 2 2D 1 2 2E 1 2 2&gt;&gt;&gt; &gt;&gt;&gt; obj.groupby(obj_dict, axis=1).sum() C++ Java PythonA 1 16 5B 7 12 10C 5 3 17D 8 5 6E 2 10 12 【03x03x02】函数分组通过函数进行分组： 123456789101112131415161718192021222324&gt;&gt;&gt; import pandas as pd&gt;&gt;&gt; import numpy as np&gt;&gt;&gt; obj = pd.DataFrame(np.random.randint(1, 10, (5,5)), columns=['a', 'b', 'c', 'd', 'e'], index=['AA', 'BBB', 'CC', 'D', 'EE'])&gt;&gt;&gt; obj a b c d eAA 3 9 5 8 2BBB 1 4 2 2 6CC 9 2 4 7 6D 2 5 5 7 1EE 8 8 8 2 2&gt;&gt;&gt; &gt;&gt;&gt; def group_key(idx): """ idx 为列索引或行索引 """ return len(idx)&gt;&gt;&gt; obj.groupby(group_key).size() # 等价于 obj.groupby(len).size()1 12 33 1dtype: int64 【03x03x03】索引层级分组通过不同索引层级进行分组： 1234567891011121314151617181920212223242526272829&gt;&gt;&gt; import pandas as pd&gt;&gt;&gt; import numpy as np&gt;&gt;&gt; columns = pd.MultiIndex.from_arrays([['Python', 'Java', 'Python', 'Java', 'Python'], ['A', 'A', 'B', 'C', 'B']], names=['language', 'index'])&gt;&gt;&gt; obj = pd.DataFrame(np.random.randint(1, 10, (5, 5)), columns=columns)&gt;&gt;&gt; objlanguage Python Java Python Java Pythonindex A A B C B0 7 1 9 8 51 4 5 4 5 62 4 3 1 9 53 6 6 3 8 14 7 9 2 8 2&gt;&gt;&gt; &gt;&gt;&gt; obj.groupby(level='language', axis=1).sum()language Java Python0 9 211 10 142 12 103 14 104 17 11&gt;&gt;&gt; &gt;&gt;&gt; obj.groupby(level='index', axis=1).sum()index A B C0 8 14 81 9 10 52 7 6 93 12 4 84 16 4 8 【03x04】分组迭代GroupBy 对象支持迭代，对于单层分组，可以产生一组二元元组，由分组名和数据块组成： 1234567891011121314151617181920212223242526272829303132333435&gt;&gt;&gt; import pandas as pd&gt;&gt;&gt; import numpy as np&gt;&gt;&gt; data = &#123;'key1' : ['a', 'b', 'a', 'b', 'a', 'b', 'a', 'a'], 'key2' : ['one', 'one', 'two', 'three', 'two', 'two', 'one', 'three'], 'data1': np.random.randn(8), 'data2': np.random.randn(8)&#125;&gt;&gt;&gt; obj = pd.DataFrame(data)&gt;&gt;&gt; obj key1 key2 data1 data20 a one -1.088762 0.6685041 b one 0.275500 0.7878442 a two -0.108417 -0.4912963 b three 0.019524 -0.3633904 a two 0.453612 0.7969995 b two 1.982858 1.5018776 a one 1.101132 -1.9283627 a three 0.524775 -1.205842&gt;&gt;&gt; &gt;&gt;&gt; for group_name, group_data in obj.groupby('key1'): print(group_name) print(group_data) a key1 key2 data1 data20 a one -1.088762 0.6685042 a two -0.108417 -0.4912964 a two 0.453612 0.7969996 a one 1.101132 -1.9283627 a three 0.524775 -1.205842b key1 key2 data1 data21 b one 0.275500 0.7878443 b three 0.019524 -0.3633905 b two 1.982858 1.501877 对于多层分组，元组的第一个元素将会是由键值组成的元组，第二个元素为数据块： 12345678910111213141516171819202122232425262728293031323334353637383940414243&gt;&gt;&gt; import pandas as pd&gt;&gt;&gt; import numpy as np&gt;&gt;&gt; data = &#123;'key1' : ['a', 'b', 'a', 'b', 'a', 'b', 'a', 'a'], 'key2' : ['one', 'one', 'two', 'three', 'two', 'two', 'one', 'three'], 'data1': np.random.randn(8), 'data2': np.random.randn(8)&#125;&gt;&gt;&gt; obj = pd.DataFrame(data)&gt;&gt;&gt; obj key1 key2 data1 data20 a one -1.088762 0.6685041 b one 0.275500 0.7878442 a two -0.108417 -0.4912963 b three 0.019524 -0.3633904 a two 0.453612 0.7969995 b two 1.982858 1.5018776 a one 1.101132 -1.9283627 a three 0.524775 -1.205842&gt;&gt;&gt; &gt;&gt;&gt; for group_name, group_data in obj.groupby(['key1', 'key2']): print(group_name) print(group_data) ('a', 'one') key1 key2 data1 data20 a one -1.088762 0.6685046 a one 1.101132 -1.928362('a', 'three') key1 key2 data1 data27 a three 0.524775 -1.205842('a', 'two') key1 key2 data1 data22 a two -0.108417 -0.4912964 a two 0.453612 0.796999('b', 'one') key1 key2 data1 data21 b one 0.2755 0.787844('b', 'three') key1 key2 data1 data23 b three 0.019524 -0.36339('b', 'two') key1 key2 data1 data25 b two 1.982858 1.501877 【03x05】对象转换GroupBy 对象支持转换成列表或字典： 123456789101112131415161718192021222324252627282930313233343536373839404142&gt;&gt;&gt; import pandas as pd&gt;&gt;&gt; import numpy as np&gt;&gt;&gt; data = &#123;'key1' : ['a', 'b', 'a', 'b', 'a', 'b', 'a', 'a'], 'key2' : ['one', 'one', 'two', 'three', 'two', 'two', 'one', 'three'], 'data1': np.random.randn(8), 'data2': np.random.randn(8)&#125;&gt;&gt;&gt; obj = pd.DataFrame(data)&gt;&gt;&gt; obj key1 key2 data1 data20 a one -0.607009 1.9483011 b one 0.150818 -0.0250952 a two -2.086024 0.3581643 b three 0.446061 1.7087974 a two 0.745457 -0.9809485 b two 0.981877 2.1593276 a one 0.804480 -0.4996617 a three 0.112884 0.004367&gt;&gt;&gt; &gt;&gt;&gt; grouped = obj.groupby('key1')&gt;&gt;&gt; list(grouped1)[('a', key1 key2 data1 data20 a one -0.607009 1.9483012 a two -2.086024 0.3581644 a two 0.745457 -0.9809486 a one 0.804480 -0.4996617 a three 0.112884 0.004367),('b', key1 key2 data1 data21 b one 0.150818 -0.0250953 b three 0.446061 1.7087975 b two 0.981877 2.159327)]&gt;&gt;&gt; &gt;&gt;&gt; dict(list(grouped1))&#123;'a': key1 key2 data1 data20 a one -0.607009 1.9483012 a two -2.086024 0.3581644 a two 0.745457 -0.9809486 a one 0.804480 -0.4996617 a three 0.112884 0.004367,'b': key1 key2 data1 data21 b one 0.150818 -0.0250953 b three 0.446061 1.7087975 b two 0.981877 2.159327&#125; 【04x00】GroupBy Apply 数据应用聚合指的是任何能够从数组产生标量值的数据转换过程，常用于对分组后的数据进行计算 【04x01】聚合函数之前的例子已经用过一些内置的聚合函数，比如 mean、count、min 以及 sum 等。常见的聚合运算如下表所示： 官方文档：https://pandas.pydata.org/docs/reference/groupby.html 方法 描述 count 非NA值的数量 describe 针对Series或各DataFrame列计算汇总统计 min 计算最小值 max 计算最大值 argmin 计算能够获取到最小值的索引位置（整数） argmax 计算能够获取到最大值的索引位置（整数） idxmin 计算能够获取到最小值的索引值 idxmax 计算能够获取到最大值的索引值 quantile 计算样本的分位数（0到1） sum 值的总和 mean 值的平均数 median 值的算术中位数（50%分位数） mad 根据平均值计算平均绝对离差 var 样本值的方差 std 样本值的标准差 应用示例： 1234567891011121314151617181920212223242526272829303132333435363738394041424344454647484950515253545556575859606162&gt;&gt;&gt; import pandas as pd&gt;&gt;&gt; import numpy as np&gt;&gt;&gt; obj = &#123;'key1' : ['a', 'b', 'a', 'b', 'a', 'b', 'a', 'a'], 'key2' : ['one', 'one', 'two', 'three', 'two', 'two', 'one', 'three'], 'data1': np.random.randint(1,10, 8), 'data2': np.random.randint(1,10, 8)&#125;&gt;&gt;&gt; obj = pd.DataFrame(obj)&gt;&gt;&gt; obj key1 key2 data1 data20 a one 9 71 b one 5 92 a two 2 43 b three 3 44 a two 5 15 b two 5 96 a one 1 87 a three 2 4&gt;&gt;&gt; &gt;&gt;&gt; obj.groupby('key1').sum() data1 data2key1 a 19 24b 13 22&gt;&gt;&gt; &gt;&gt;&gt; obj.groupby('key1').max() key2 data1 data2key1 a two 9 8b two 5 9&gt;&gt;&gt; &gt;&gt;&gt; obj.groupby('key1').min() key2 data1 data2key1 a one 1 1b one 3 4&gt;&gt;&gt; &gt;&gt;&gt; obj.groupby('key1').mean() data1 data2key1 a 3.800000 4.800000b 4.333333 7.333333&gt;&gt;&gt; &gt;&gt;&gt; obj.groupby('key1').size()key1a 5b 3dtype: int64&gt;&gt;&gt; &gt;&gt;&gt; obj.groupby('key1').count() key2 data1 data2key1 a 5 5 5b 3 3 3&gt;&gt;&gt; &gt;&gt;&gt; obj.groupby('key1').describe() data1 ... data2 count mean std min 25% ... min 25% 50% 75% maxkey1 ... a 5.0 3.800000 3.271085 1.0 2.0 ... 1.0 4.0 4.0 7.0 8.0b 3.0 4.333333 1.154701 3.0 4.0 ... 4.0 6.5 9.0 9.0 9.0[2 rows x 16 columns] 【04x02】自定义函数如果自带的内置函数满足不了我们的要求，则可以自定义一个聚合函数，然后传入 GroupBy.agg(func) 或 GroupBy.aggregate(func) 方法中即可。func 的参数为 groupby 索引对应的记录。 123456789101112131415161718192021222324252627282930313233&gt;&gt;&gt; import pandas as pd&gt;&gt;&gt; import numpy as np&gt;&gt;&gt; obj = &#123;'key1' : ['a', 'b', 'a', 'b', 'a', 'b', 'a', 'a'], 'key2' : ['one', 'one', 'two', 'three', 'two', 'two', 'one', 'three'], 'data1': np.random.randint(1,10, 8), 'data2': np.random.randint(1,10, 8)&#125;&gt;&gt;&gt; obj = pd.DataFrame(obj)&gt;&gt;&gt; obj key1 key2 data1 data20 a one 9 71 b one 5 92 a two 2 43 b three 3 44 a two 5 15 b two 5 96 a one 1 87 a three 2 4&gt;&gt;&gt; &gt;&gt;&gt; def peak_range(df): return df.max() - df.min()&gt;&gt;&gt; &gt;&gt;&gt; obj.groupby('key1').agg(peak_range) data1 data2key1 a 8 7b 2 5&gt;&gt;&gt; &gt;&gt;&gt; obj.groupby('key1').agg(lambda df : df.max() - df.min()) data1 data2key1 a 8 7b 2 5 【04x03】对不同列作用不同函数使用字典可以对不同列作用不同的聚合函数： 123456789101112131415161718192021222324252627282930313233&gt;&gt;&gt; import pandas as pd&gt;&gt;&gt; import numpy as np&gt;&gt;&gt; obj = &#123;'key1' : ['a', 'b', 'a', 'b', 'a', 'b', 'a', 'a'], 'key2' : ['one', 'one', 'two', 'three', 'two', 'two', 'one', 'three'], 'data1': np.random.randint(1,10, 8), 'data2': np.random.randint(1,10, 8)&#125;&gt;&gt;&gt; obj = pd.DataFrame(obj)&gt;&gt;&gt; obj key1 key2 data1 data20 a one 9 71 b one 5 92 a two 2 43 b three 3 44 a two 5 15 b two 5 96 a one 1 87 a three 2 4&gt;&gt;&gt; &gt;&gt;&gt; dict1 = &#123;'data1':'mean', 'data2':'sum'&#125;&gt;&gt;&gt; dict2 = &#123;'data1':['mean','max'], 'data2':'sum'&#125;&gt;&gt;&gt; &gt;&gt;&gt; obj.groupby('key1').agg(dict1) data1 data2key1 a 3.800000 24b 4.333333 22&gt;&gt;&gt; &gt;&gt;&gt; obj.groupby('key1').agg(dict2) data1 data2 mean max sumkey1 a 3.800000 9 24b 4.333333 5 22 【04x04】GroupBy.apply()apply() 方法会将待处理的对象拆分成多个片段，然后对各片段调用传入的函数，最后尝试将各片段组合到一起。 123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960&gt;&gt;&gt; import pandas as pd&gt;&gt;&gt; obj = pd.DataFrame(&#123;'A':['bob','sos','bob','sos','bob','sos','bob','bob'], 'B':['one','one','two','three','two','two','one','three'], 'C':[3,1,4,1,5,9,2,6], 'D':[1,2,3,4,5,6,7,8]&#125;)&gt;&gt;&gt; obj A B C D0 bob one 3 11 sos one 1 22 bob two 4 33 sos three 1 44 bob two 5 55 sos two 9 66 bob one 2 77 bob three 6 8&gt;&gt;&gt; &gt;&gt;&gt; grouped = obj.groupby('A')&gt;&gt;&gt; for name, group in grouped: print(name) print(group) bob A B C D0 bob one 3 12 bob two 4 34 bob two 5 56 bob one 2 77 bob three 6 8sos A B C D1 sos one 1 23 sos three 1 45 sos two 9 6&gt;&gt;&gt; &gt;&gt;&gt; grouped.apply(lambda x:x.describe()) # 对 bob 和 sos 两组数据使用 describe 方法 C DA bob count 5.000000 5.000000 mean 4.000000 4.800000 std 1.581139 2.863564 min 2.000000 1.000000 25% 3.000000 3.000000 50% 4.000000 5.000000 75% 5.000000 7.000000 max 6.000000 8.000000sos count 3.000000 3.000000 mean 3.666667 4.000000 std 4.618802 2.000000 min 1.000000 2.000000 25% 1.000000 3.000000 50% 1.000000 4.000000 75% 5.000000 5.000000 max 9.000000 6.000000&gt;&gt;&gt;&gt;&gt;&gt; grouped.apply(lambda x:x.min()) # # 对 bob 和 sos 两组数据使用 min 方法 A B C DA bob bob one 2 1sos sos one 1 2 12345这里是一段防爬虫文本，请读者忽略。本文原创首发于 CSDN，作者 TRHX。博客首页：https://itrhx.blog.csdn.net/本文链接：https://itrhx.blog.csdn.net/article/details/106804881未经授权，禁止转载！恶意转载，后果自负！尊重原创，远离剽窃！]]></content>
<content type="text"><![CDATA[Pandas 系列文章： Python 数据分析三剑客之 Pandas（一）：认识 Pandas 及其 Series、DataFrame 对象 Python 数据分析三剑客之 Pandas（二）：Index 索引对象以及各种索引操作 Python 数据分析三剑客之 Pandas（三）：算术运算与缺失值的处理 Python 数据分析三剑客之 Pandas（四）：函数应用、映射、排序和层级索引 Python 数据分析三剑客之 Pandas（五）：统计计算与统计描述 Python 数据分析三剑客之 Pandas（六）：GroupBy 数据分裂、应用与合并 Python 数据分析三剑客之 Pandas（七）：合并数据集 Python 数据分析三剑客之 Pandas（八）：数据重塑、重复数据处理与数据替换 Python 数据分析三剑客之 Pandas（九）：时间序列 Python 数据分析三剑客之 Pandas（十）：数据读写 专栏： 【NumPy 专栏】【Pandas 专栏】【Matplotlib 专栏】 推荐学习资料与网站： 【NumPy 中文网】【Pandas 中文网】【Matplotlib 中文网】【NumPy、Matplotlib、Pandas 速查表】 12345这里是一段防爬虫文本，请读者忽略。本文原创首发于 CSDN，作者 TRHX。博客首页：https://itrhx.blog.csdn.net/本文链接：https://itrhx.blog.csdn.net/article/details/106804881未经授权，禁止转载！恶意转载，后果自负！尊重原创，远离剽窃！ 【01x00】GroupBy 机制对数据集进行分组并对各组应用一个函数（无论是聚合还是转换），通常是数据分析工作中的重要环节。在将数据集加载、融合、准备好之后，通常就是计算分组统计或生成透视表。Pandas 提供了一个灵活高效的 GroupBy 功能，虽然“分组”（group by）这个名字是借用 SQL 数据库语言的命令，但其理念引用发明 R 语言 frame 的 Hadley Wickham 的观点可能更合适：分裂（Split）、应用（Apply）和组合（Combine）。 分组运算过程：Split —&gt; Apply —&gt; Combine 分裂（Split）：根据某些标准将数据分组； 应用（Apply）：对每个组独立应用一个函数； 合并（Combine）：把每个分组的计算结果合并起来。 官方介绍：https://pandas.pydata.org/docs/user_guide/groupby.html 【02x00】GroupBy 对象常见的 GroupBy 对象：Series.groupby、DataFrame.groupby，基本语法如下： 123456789Series.groupby(self, by=None, axis=0, level=None, as_index: bool = True, sort: bool = True, group_keys: bool = True, squeeze: bool = False, observed: bool = False) → ’groupby_generic.SeriesGroupBy’ 123456789DataFrame.groupby(self, by=None, axis=0, level=None, as_index: bool = True, sort: bool = True, group_keys: bool = True, squeeze: bool = False, observed: bool = False) → ’groupby_generic.DataFrameGroupBy’ 官方文档： https://pandas.pydata.org/docs/reference/api/pandas.Series.groupby.html https://pandas.pydata.org/docs/reference/api/pandas.DataFrame.groupby.html 常用参数解释如下： 参数 描述 by 映射、函数、标签或标签列表，用于确定分组依据的分组。如果 by 是函数，则会在对象索引的每个值上调用它。 如果传递了 dict 或 Series，则将使用 Series 或 dict 的值来确定组（将 Series 的值首先对齐；请参见.align() 方法）。 如果传递了 ndarray，则按原样使用这些值来确定组。标签或标签列表可以按自身中的列传递给分组。 注意，元组被解释为（单个）键 axis 沿指定轴拆分，默认 0，0 or ‘index’，1 or ‘columns’，只有在 DataFrame 中才有 1 or &#39;columns’ level 如果轴是 MultiIndex（层次结构），则按特定层级进行分组，默认 None as_index bool 类型，默认 True，对于聚合输出，返回以组标签为索引的对象。仅与 DataFrame 输入相关。as_index=False 实际上是“SQL样式”分组输出 sort bool 类型，默认 True，对组键排序。关闭此选项可获得更好的性能。注：这不影响每组的观察顺序。Groupby 保留每个组中行的顺序 group_keys bool 类型，默认 True，调用 apply 方法时，是否将组键（keys）添加到索引（ index）以标识块 squeeze bool 类型，默认 False，如果可能，减少返回类型的维度，否则返回一致的类型 groupby() 进行分组，GroupBy 对象没有进行实际运算，只是包含分组的中间数据，示例如下： 123456789101112131415161718192021222324&gt;&gt;&gt; import pandas as pd&gt;&gt;&gt; import numpy as np&gt;&gt;&gt; data = &#123;'key1' : ['a', 'b', 'a', 'b', 'a', 'b', 'a', 'a'], 'key2' : ['one', 'one', 'two', 'three', 'two', 'two', 'one', 'three'], 'data1': np.random.randn(8), 'data2': np.random.randn(8)&#125;&gt;&gt;&gt; &gt;&gt;&gt; obj = pd.DataFrame(data)&gt;&gt;&gt; obj key1 key2 data1 data20 a one -0.804160 -0.8689051 b one -0.086990 0.3257412 a two 0.757992 0.5411013 b three -0.281435 0.0978414 a two 0.817757 -0.6436995 b two -0.462760 -0.3211966 a one -0.403699 0.6021387 a three 0.883940 -0.850526&gt;&gt;&gt; &gt;&gt;&gt; obj.groupby('key1')&lt;pandas.core.groupby.generic.DataFrameGroupBy object at 0x03CDB7C0&gt;&gt;&gt;&gt; &gt;&gt;&gt; obj['data1'].groupby(obj['key1'])&lt;pandas.core.groupby.generic.SeriesGroupBy object at 0x03CDB748&gt; 【03x00】GroupBy Split 数据分裂【03x01】分组运算前面通过 groupby() 方法获得了一个 GroupBy 对象，它实际上还没有进行任何计算，只是含有一些有关分组键 obj[&#39;key1&#39;] 的中间数据而已。换句话说，该对象已经有了接下来对各分组执行运算所需的一切信息。例如，我们可以调用 GroupBy 的 mean() 方法来计算分组平均值，size() 方法返回每个分组的元素个数： 123456789101112131415161718192021222324252627282930313233343536373839404142434445&gt;&gt;&gt; import pandas as pd&gt;&gt;&gt; import numpy as np&gt;&gt;&gt; data = &#123;'key1' : ['a', 'b', 'a', 'b', 'a', 'b', 'a', 'a'], 'key2' : ['one', 'one', 'two', 'three', 'two', 'two', 'one', 'three'], 'data1': np.random.randn(8), 'data2': np.random.randn(8)&#125;&gt;&gt;&gt; &gt;&gt;&gt; obj = pd.DataFrame(data)&gt;&gt;&gt; obj key1 key2 data1 data20 a one -0.544099 -0.6140791 b one 2.193712 0.1010052 a two -0.004683 0.8827703 b three 0.312858 1.7321054 a two 0.011089 0.0895875 b two 0.292165 1.3276386 a one -1.433291 -0.2389717 a three -0.004724 -2.117326&gt;&gt;&gt; &gt;&gt;&gt; grouped1 = obj.groupby('key1')&gt;&gt;&gt; grouped2 = obj['data1'].groupby(obj['key1'])&gt;&gt;&gt; &gt;&gt;&gt; grouped1.mean() data1 data2key1 a -0.395142 -0.399604b 0.932912 1.053583&gt;&gt;&gt; &gt;&gt;&gt; grouped2.mean()key1a -0.395142b 0.932912Name: data1, dtype: float64&gt;&gt;&gt;&gt;&gt;&gt; grouped1.size()key1a 5b 3dtype: int64&gt;&gt;&gt; &gt;&gt;&gt; grouped2.size()key1a 5b 3Name: data1, dtype: int64 【03x02】按类型按列分组groupby() 方法 axis 参数默认是 0，通过设置也可以在其他任何轴上进行分组，也支持按照类型（dtype）进行分组： 12345678910111213141516171819202122232425262728293031323334353637383940&gt;&gt;&gt; import pandas as pd&gt;&gt;&gt; import numpy as np&gt;&gt;&gt; data = &#123;'key1' : ['a', 'b', 'a', 'b', 'a', 'b', 'a', 'a'], 'key2' : ['one', 'one', 'two', 'three', 'two', 'two', 'one', 'three'], 'data1': np.random.randn(8), 'data2': np.random.randn(8)&#125;&gt;&gt;&gt; obj = pd.DataFrame(data)&gt;&gt;&gt; obj key1 key2 data1 data20 a one -0.607009 1.9483011 b one 0.150818 -0.0250952 a two -2.086024 0.3581643 b three 0.446061 1.7087974 a two 0.745457 -0.9809485 b two 0.981877 2.1593276 a one 0.804480 -0.4996617 a three 0.112884 0.004367&gt;&gt;&gt; &gt;&gt;&gt; obj.dtypeskey1 objectkey2 objectdata1 float64data2 float64dtype: object&gt;&gt;&gt; &gt;&gt;&gt; obj.groupby(obj.dtypes, axis=1).size()float64 2object 2dtype: int64&gt;&gt;&gt; &gt;&gt;&gt; obj.groupby(obj.dtypes, axis=1).sum() float64 object0 1.341291 aone1 0.125723 bone2 -1.727860 atwo3 2.154858 bthree4 -0.235491 atwo5 3.141203 btwo6 0.304819 aone7 0.117251 athree 【03x03】自定义分组groupby() 方法中可以一次传入多个数组的列表，也可以自定义一组分组键。也可以通过一个字典、一个函数，或者按照索引层级进行分组。 传入多个数组的列表： 12345678910111213141516171819202122232425262728293031323334&gt;&gt;&gt; import pandas as pd&gt;&gt;&gt; import numpy as np&gt;&gt;&gt; data = &#123;'key1' : ['a', 'b', 'a', 'b', 'a', 'b', 'a', 'a'], 'key2' : ['one', 'one', 'two', 'three', 'two', 'two', 'one', 'three'], 'data1': np.random.randn(8), 'data2': np.random.randn(8)&#125;&gt;&gt;&gt; obj = pd.DataFrame(data)&gt;&gt;&gt; obj key1 key2 data1 data20 a one -0.841652 0.6880551 b one 0.510042 -0.5611712 a two -0.418862 -0.1459833 b three -1.104698 0.5631584 a two 0.329527 -0.8931085 b two 0.753653 -0.3425206 a one -0.882527 -1.1213297 a three 1.726794 0.160244&gt;&gt;&gt; &gt;&gt;&gt; means = obj['data1'].groupby([obj['key1'], obj['key2']]).mean()&gt;&gt;&gt; meanskey1 key2 a one -0.862090 three 1.726794 two -0.044667b one 0.510042 three -1.104698 two 0.753653Name: data1, dtype: float64&gt;&gt;&gt; &gt;&gt;&gt; means.unstack()key2 one three twokey1 a -0.862090 1.726794 -0.044667b 0.510042 -1.104698 0.753653 自定义分组键： 1234567891011121314151617181920212223&gt;&gt;&gt; import pandas as pd&gt;&gt;&gt; import numpy as np&gt;&gt;&gt; obj = pd.DataFrame(&#123;'key1' : ['a', 'a', 'b', 'b', 'a'], 'key2' : ['one', 'two', 'one', 'two', 'one'], 'data1' : np.random.randn(5), 'data2' : np.random.randn(5)&#125;)&gt;&gt;&gt; obj key1 key2 data1 data20 a one -0.024003 0.3504801 a two -0.767534 -0.1004262 b one -0.594983 -1.9455803 b two -0.374482 0.8175924 a one 0.755452 -0.137759&gt;&gt;&gt; &gt;&gt;&gt; states = np.array(['Wuhan', 'Beijing', 'Beijing', 'Wuhan', 'Wuhan'])&gt;&gt;&gt; years = np.array([2005, 2005, 2006, 2005, 2006])&gt;&gt;&gt; &gt;&gt;&gt; obj['data1'].groupby([states, years]).mean()Beijing 2005 -0.767534 2006 -0.594983Wuhan 2005 -0.199242 2006 0.755452Name: data1, dtype: float64 【03x03x01】字典分组通过字典进行分组： 1234567891011121314151617181920212223242526272829303132333435&gt;&gt;&gt; import pandas as pd&gt;&gt;&gt; import numpy as np&gt;&gt;&gt; obj = pd.DataFrame(np.random.randint(1, 10, (5,5)), columns=['a', 'b', 'c', 'd', 'e'], index=['A', 'B', 'C', 'D', 'E'])&gt;&gt;&gt; obj a b c d eA 1 4 7 1 9B 8 2 4 7 8C 9 8 2 5 1D 2 4 2 8 3E 7 5 7 2 3&gt;&gt;&gt; &gt;&gt;&gt; obj_dict = &#123;'a':'Python', 'b':'Python', 'c':'Java', 'd':'C++', 'e':'Java'&#125;&gt;&gt;&gt; obj.groupby(obj_dict, axis=1).size()C++ 1Java 2Python 2dtype: int64&gt;&gt;&gt; &gt;&gt;&gt; obj.groupby(obj_dict, axis=1).count() C++ Java PythonA 1 2 2B 1 2 2C 1 2 2D 1 2 2E 1 2 2&gt;&gt;&gt; &gt;&gt;&gt; obj.groupby(obj_dict, axis=1).sum() C++ Java PythonA 1 16 5B 7 12 10C 5 3 17D 8 5 6E 2 10 12 【03x03x02】函数分组通过函数进行分组： 123456789101112131415161718192021222324&gt;&gt;&gt; import pandas as pd&gt;&gt;&gt; import numpy as np&gt;&gt;&gt; obj = pd.DataFrame(np.random.randint(1, 10, (5,5)), columns=['a', 'b', 'c', 'd', 'e'], index=['AA', 'BBB', 'CC', 'D', 'EE'])&gt;&gt;&gt; obj a b c d eAA 3 9 5 8 2BBB 1 4 2 2 6CC 9 2 4 7 6D 2 5 5 7 1EE 8 8 8 2 2&gt;&gt;&gt; &gt;&gt;&gt; def group_key(idx): """ idx 为列索引或行索引 """ return len(idx)&gt;&gt;&gt; obj.groupby(group_key).size() # 等价于 obj.groupby(len).size()1 12 33 1dtype: int64 【03x03x03】索引层级分组通过不同索引层级进行分组： 1234567891011121314151617181920212223242526272829&gt;&gt;&gt; import pandas as pd&gt;&gt;&gt; import numpy as np&gt;&gt;&gt; columns = pd.MultiIndex.from_arrays([['Python', 'Java', 'Python', 'Java', 'Python'], ['A', 'A', 'B', 'C', 'B']], names=['language', 'index'])&gt;&gt;&gt; obj = pd.DataFrame(np.random.randint(1, 10, (5, 5)), columns=columns)&gt;&gt;&gt; objlanguage Python Java Python Java Pythonindex A A B C B0 7 1 9 8 51 4 5 4 5 62 4 3 1 9 53 6 6 3 8 14 7 9 2 8 2&gt;&gt;&gt; &gt;&gt;&gt; obj.groupby(level='language', axis=1).sum()language Java Python0 9 211 10 142 12 103 14 104 17 11&gt;&gt;&gt; &gt;&gt;&gt; obj.groupby(level='index', axis=1).sum()index A B C0 8 14 81 9 10 52 7 6 93 12 4 84 16 4 8 【03x04】分组迭代GroupBy 对象支持迭代，对于单层分组，可以产生一组二元元组，由分组名和数据块组成： 1234567891011121314151617181920212223242526272829303132333435&gt;&gt;&gt; import pandas as pd&gt;&gt;&gt; import numpy as np&gt;&gt;&gt; data = &#123;'key1' : ['a', 'b', 'a', 'b', 'a', 'b', 'a', 'a'], 'key2' : ['one', 'one', 'two', 'three', 'two', 'two', 'one', 'three'], 'data1': np.random.randn(8), 'data2': np.random.randn(8)&#125;&gt;&gt;&gt; obj = pd.DataFrame(data)&gt;&gt;&gt; obj key1 key2 data1 data20 a one -1.088762 0.6685041 b one 0.275500 0.7878442 a two -0.108417 -0.4912963 b three 0.019524 -0.3633904 a two 0.453612 0.7969995 b two 1.982858 1.5018776 a one 1.101132 -1.9283627 a three 0.524775 -1.205842&gt;&gt;&gt; &gt;&gt;&gt; for group_name, group_data in obj.groupby('key1'): print(group_name) print(group_data) a key1 key2 data1 data20 a one -1.088762 0.6685042 a two -0.108417 -0.4912964 a two 0.453612 0.7969996 a one 1.101132 -1.9283627 a three 0.524775 -1.205842b key1 key2 data1 data21 b one 0.275500 0.7878443 b three 0.019524 -0.3633905 b two 1.982858 1.501877 对于多层分组，元组的第一个元素将会是由键值组成的元组，第二个元素为数据块： 12345678910111213141516171819202122232425262728293031323334353637383940414243&gt;&gt;&gt; import pandas as pd&gt;&gt;&gt; import numpy as np&gt;&gt;&gt; data = &#123;'key1' : ['a', 'b', 'a', 'b', 'a', 'b', 'a', 'a'], 'key2' : ['one', 'one', 'two', 'three', 'two', 'two', 'one', 'three'], 'data1': np.random.randn(8), 'data2': np.random.randn(8)&#125;&gt;&gt;&gt; obj = pd.DataFrame(data)&gt;&gt;&gt; obj key1 key2 data1 data20 a one -1.088762 0.6685041 b one 0.275500 0.7878442 a two -0.108417 -0.4912963 b three 0.019524 -0.3633904 a two 0.453612 0.7969995 b two 1.982858 1.5018776 a one 1.101132 -1.9283627 a three 0.524775 -1.205842&gt;&gt;&gt; &gt;&gt;&gt; for group_name, group_data in obj.groupby(['key1', 'key2']): print(group_name) print(group_data) ('a', 'one') key1 key2 data1 data20 a one -1.088762 0.6685046 a one 1.101132 -1.928362('a', 'three') key1 key2 data1 data27 a three 0.524775 -1.205842('a', 'two') key1 key2 data1 data22 a two -0.108417 -0.4912964 a two 0.453612 0.796999('b', 'one') key1 key2 data1 data21 b one 0.2755 0.787844('b', 'three') key1 key2 data1 data23 b three 0.019524 -0.36339('b', 'two') key1 key2 data1 data25 b two 1.982858 1.501877 【03x05】对象转换GroupBy 对象支持转换成列表或字典： 123456789101112131415161718192021222324252627282930313233343536373839404142&gt;&gt;&gt; import pandas as pd&gt;&gt;&gt; import numpy as np&gt;&gt;&gt; data = &#123;'key1' : ['a', 'b', 'a', 'b', 'a', 'b', 'a', 'a'], 'key2' : ['one', 'one', 'two', 'three', 'two', 'two', 'one', 'three'], 'data1': np.random.randn(8), 'data2': np.random.randn(8)&#125;&gt;&gt;&gt; obj = pd.DataFrame(data)&gt;&gt;&gt; obj key1 key2 data1 data20 a one -0.607009 1.9483011 b one 0.150818 -0.0250952 a two -2.086024 0.3581643 b three 0.446061 1.7087974 a two 0.745457 -0.9809485 b two 0.981877 2.1593276 a one 0.804480 -0.4996617 a three 0.112884 0.004367&gt;&gt;&gt; &gt;&gt;&gt; grouped = obj.groupby('key1')&gt;&gt;&gt; list(grouped)[('a', key1 key2 data1 data20 a one -0.607009 1.9483012 a two -2.086024 0.3581644 a two 0.745457 -0.9809486 a one 0.804480 -0.4996617 a three 0.112884 0.004367),('b', key1 key2 data1 data21 b one 0.150818 -0.0250953 b three 0.446061 1.7087975 b two 0.981877 2.159327)]&gt;&gt;&gt;&gt;&gt;&gt; dict(list(grouped))&#123;'a': key1 key2 data1 data20 a one -0.607009 1.9483012 a two -2.086024 0.3581644 a two 0.745457 -0.9809486 a one 0.804480 -0.4996617 a three 0.112884 0.004367,'b': key1 key2 data1 data21 b one 0.150818 -0.0250953 b three 0.446061 1.7087975 b two 0.981877 2.159327&#125; 【04x00】GroupBy Apply 数据应用聚合指的是任何能够从数组产生标量值的数据转换过程，常用于对分组后的数据进行计算 【04x01】聚合函数之前的例子已经用过一些内置的聚合函数，比如 mean、count、min 以及 sum 等。常见的聚合运算如下表所示： 官方文档：https://pandas.pydata.org/docs/reference/groupby.html 方法 描述 count 非NA值的数量 describe 针对Series或各DataFrame列计算汇总统计 min 计算最小值 max 计算最大值 argmin 计算能够获取到最小值的索引位置（整数） argmax 计算能够获取到最大值的索引位置（整数） idxmin 计算能够获取到最小值的索引值 idxmax 计算能够获取到最大值的索引值 quantile 计算样本的分位数（0到1） sum 值的总和 mean 值的平均数 median 值的算术中位数（50%分位数） mad 根据平均值计算平均绝对离差 var 样本值的方差 std 样本值的标准差 应用示例： 1234567891011121314151617181920212223242526272829303132333435363738394041424344454647484950515253545556575859606162&gt;&gt;&gt; import pandas as pd&gt;&gt;&gt; import numpy as np&gt;&gt;&gt; obj = &#123;'key1' : ['a', 'b', 'a', 'b', 'a', 'b', 'a', 'a'], 'key2' : ['one', 'one', 'two', 'three', 'two', 'two', 'one', 'three'], 'data1': np.random.randint(1,10, 8), 'data2': np.random.randint(1,10, 8)&#125;&gt;&gt;&gt; obj = pd.DataFrame(obj)&gt;&gt;&gt; obj key1 key2 data1 data20 a one 9 71 b one 5 92 a two 2 43 b three 3 44 a two 5 15 b two 5 96 a one 1 87 a three 2 4&gt;&gt;&gt; &gt;&gt;&gt; obj.groupby('key1').sum() data1 data2key1 a 19 24b 13 22&gt;&gt;&gt; &gt;&gt;&gt; obj.groupby('key1').max() key2 data1 data2key1 a two 9 8b two 5 9&gt;&gt;&gt; &gt;&gt;&gt; obj.groupby('key1').min() key2 data1 data2key1 a one 1 1b one 3 4&gt;&gt;&gt; &gt;&gt;&gt; obj.groupby('key1').mean() data1 data2key1 a 3.800000 4.800000b 4.333333 7.333333&gt;&gt;&gt; &gt;&gt;&gt; obj.groupby('key1').size()key1a 5b 3dtype: int64&gt;&gt;&gt; &gt;&gt;&gt; obj.groupby('key1').count() key2 data1 data2key1 a 5 5 5b 3 3 3&gt;&gt;&gt; &gt;&gt;&gt; obj.groupby('key1').describe() data1 ... data2 count mean std min 25% ... min 25% 50% 75% maxkey1 ... a 5.0 3.800000 3.271085 1.0 2.0 ... 1.0 4.0 4.0 7.0 8.0b 3.0 4.333333 1.154701 3.0 4.0 ... 4.0 6.5 9.0 9.0 9.0[2 rows x 16 columns] 【04x02】自定义函数如果自带的内置函数满足不了我们的要求，则可以自定义一个聚合函数，然后传入 GroupBy.agg(func) 或 GroupBy.aggregate(func) 方法中即可。func 的参数为 groupby 索引对应的记录。 123456789101112131415161718192021222324252627282930313233&gt;&gt;&gt; import pandas as pd&gt;&gt;&gt; import numpy as np&gt;&gt;&gt; obj = &#123;'key1' : ['a', 'b', 'a', 'b', 'a', 'b', 'a', 'a'], 'key2' : ['one', 'one', 'two', 'three', 'two', 'two', 'one', 'three'], 'data1': np.random.randint(1,10, 8), 'data2': np.random.randint(1,10, 8)&#125;&gt;&gt;&gt; obj = pd.DataFrame(obj)&gt;&gt;&gt; obj key1 key2 data1 data20 a one 9 71 b one 5 92 a two 2 43 b three 3 44 a two 5 15 b two 5 96 a one 1 87 a three 2 4&gt;&gt;&gt; &gt;&gt;&gt; def peak_range(df): return df.max() - df.min()&gt;&gt;&gt; &gt;&gt;&gt; obj.groupby('key1').agg(peak_range) data1 data2key1 a 8 7b 2 5&gt;&gt;&gt; &gt;&gt;&gt; obj.groupby('key1').agg(lambda df : df.max() - df.min()) data1 data2key1 a 8 7b 2 5 【04x03】对不同列作用不同函数使用字典可以对不同列作用不同的聚合函数： 123456789101112131415161718192021222324252627282930313233&gt;&gt;&gt; import pandas as pd&gt;&gt;&gt; import numpy as np&gt;&gt;&gt; obj = &#123;'key1' : ['a', 'b', 'a', 'b', 'a', 'b', 'a', 'a'], 'key2' : ['one', 'one', 'two', 'three', 'two', 'two', 'one', 'three'], 'data1': np.random.randint(1,10, 8), 'data2': np.random.randint(1,10, 8)&#125;&gt;&gt;&gt; obj = pd.DataFrame(obj)&gt;&gt;&gt; obj key1 key2 data1 data20 a one 9 71 b one 5 92 a two 2 43 b three 3 44 a two 5 15 b two 5 96 a one 1 87 a three 2 4&gt;&gt;&gt; &gt;&gt;&gt; dict1 = &#123;'data1':'mean', 'data2':'sum'&#125;&gt;&gt;&gt; dict2 = &#123;'data1':['mean','max'], 'data2':'sum'&#125;&gt;&gt;&gt; &gt;&gt;&gt; obj.groupby('key1').agg(dict1) data1 data2key1 a 3.800000 24b 4.333333 22&gt;&gt;&gt; &gt;&gt;&gt; obj.groupby('key1').agg(dict2) data1 data2 mean max sumkey1 a 3.800000 9 24b 4.333333 5 22 【04x04】GroupBy.apply()apply() 方法会将待处理的对象拆分成多个片段，然后对各片段调用传入的函数，最后尝试将各片段组合到一起。 123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960&gt;&gt;&gt; import pandas as pd&gt;&gt;&gt; obj = pd.DataFrame(&#123;'A':['bob','sos','bob','sos','bob','sos','bob','bob'], 'B':['one','one','two','three','two','two','one','three'], 'C':[3,1,4,1,5,9,2,6], 'D':[1,2,3,4,5,6,7,8]&#125;)&gt;&gt;&gt; obj A B C D0 bob one 3 11 sos one 1 22 bob two 4 33 sos three 1 44 bob two 5 55 sos two 9 66 bob one 2 77 bob three 6 8&gt;&gt;&gt; &gt;&gt;&gt; grouped = obj.groupby('A')&gt;&gt;&gt; for name, group in grouped: print(name) print(group) bob A B C D0 bob one 3 12 bob two 4 34 bob two 5 56 bob one 2 77 bob three 6 8sos A B C D1 sos one 1 23 sos three 1 45 sos two 9 6&gt;&gt;&gt; &gt;&gt;&gt; grouped.apply(lambda x:x.describe()) # 对 bob 和 sos 两组数据使用 describe 方法 C DA bob count 5.000000 5.000000 mean 4.000000 4.800000 std 1.581139 2.863564 min 2.000000 1.000000 25% 3.000000 3.000000 50% 4.000000 5.000000 75% 5.000000 7.000000 max 6.000000 8.000000sos count 3.000000 3.000000 mean 3.666667 4.000000 std 4.618802 2.000000 min 1.000000 2.000000 25% 1.000000 3.000000 50% 1.000000 4.000000 75% 5.000000 5.000000 max 9.000000 6.000000&gt;&gt;&gt;&gt;&gt;&gt; grouped.apply(lambda x:x.min()) # # 对 bob 和 sos 两组数据使用 min 方法 A B C DA bob bob one 2 1sos sos one 1 2 12345这里是一段防爬虫文本，请读者忽略。本文原创首发于 CSDN，作者 TRHX。博客首页：https://itrhx.blog.csdn.net/本文链接：https://itrhx.blog.csdn.net/article/details/106804881未经授权，禁止转载！恶意转载，后果自负！尊重原创，远离剽窃！]]></content>
<categories>
<category>Python 数据分析</category>
<category>Pandas</category>
......@@ -1140,8 +1140,8 @@
<category>Hexo</category>
</categories>
<tags>
<tag>Hexo</tag>
<tag>主题个性化</tag>
<tag>Hexo</tag>
<tag>Material X</tag>
<tag>spfk</tag>
</tags>
......@@ -1166,8 +1166,8 @@
<category>Hexo</category>
</categories>
<tags>
<tag>Github Pages</tag>
<tag>Hexo</tag>
<tag>Github Pages</tag>
</tags>
</entry>
<entry>
sitemap.xml
浏览文件 @ aed7b9cf
<?xml version="1.0" encoding="UTF-8"?>
<urlset xmlns="http://www.sitemaps.org/schemas/sitemap/0.9">
<url>
<loc>https://www.itrhx.com/2020/06/17/A84-Pandas-06/</loc>
<lastmod>2020-08-07T01:13:45.027Z</lastmod>
</url>
<url>
<loc>https://www.itrhx.com/2020/07/13/A90-pyspider-51job/</loc>
......@@ -22,13 +29,6 @@
</url>
<url>
<loc>https://www.itrhx.com/2020/06/17/A84-Pandas-06/</loc>
<lastmod>2020-08-06T03:31:25.195Z</lastmod>
</url>
<url>
<loc>https://www.itrhx.com/2020/06/11/A79-Pandas-01/</loc>
......@@ -422,28 +422,28 @@
</url>
<url>
<loc>https://www.itrhx.com/games/2048/index.html</loc>
<loc>https://www.itrhx.com/games/element/index.html</loc>
<lastmod>2019-12-29T06:55:50.751Z</lastmod>
</url>
<url>
<loc>https://www.itrhx.com/box/about/index.html</loc>
<loc>https://www.itrhx.com/games/cat/index.html</loc>
<lastmod>2019-12-29T06:55:50.751Z</lastmod>
</url>
<url>
<loc>https://www.itrhx.com/games/cat/index.html</loc>
<loc>https://www.itrhx.com/box/about/index.html</loc>
<lastmod>2019-12-29T06:55:50.751Z</lastmod>
</url>
<url>
<loc>https://www.itrhx.com/games/element/index.html</loc>
<loc>https://www.itrhx.com/games/2048/index.html</loc>
<lastmod>2019-12-29T06:55:50.751Z</lastmod>
......
tags/Hexo/index.html
浏览文件 @ aed7b9cf
......@@ -1342,10 +1342,10 @@
<div class="full-width auto-padding tags">
<a href="/tags/Hexo/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;Hexo</a>
<a href="/tags/主题个性化/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;主题个性化</a>
<a href="/tags/Hexo/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;Hexo</a>
<a href="/tags/Material-X/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;Material X</a>
<a href="/tags/spfk/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;spfk</a>
......@@ -1492,10 +1492,10 @@
<div class="full-width auto-padding tags">
<a href="/tags/Github-Pages/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;Github Pages</a>
<a href="/tags/Hexo/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;Hexo</a>
<a href="/tags/Github-Pages/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;Github Pages</a>
</div>
</section>
......
tags/Material-X/index.html
浏览文件 @ aed7b9cf
......@@ -547,10 +547,10 @@
<div class="full-width auto-padding tags">
<a href="/tags/Hexo/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;Hexo</a>
<a href="/tags/主题个性化/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;主题个性化</a>
<a href="/tags/Hexo/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;Hexo</a>
<a href="/tags/Material-X/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;Material X</a>
<a href="/tags/spfk/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;spfk</a>
......
tags/spfk/index.html
浏览文件 @ aed7b9cf
......@@ -547,10 +547,10 @@
<div class="full-width auto-padding tags">
<a href="/tags/Hexo/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;Hexo</a>
<a href="/tags/主题个性化/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;主题个性化</a>
<a href="/tags/Hexo/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;Hexo</a>
<a href="/tags/Material-X/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;Material X</a>
<a href="/tags/spfk/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;spfk</a>
......
tags/主题个性化/index.html
浏览文件 @ aed7b9cf
......@@ -547,10 +547,10 @@
<div class="full-width auto-padding tags">
<a href="/tags/Hexo/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;Hexo</a>
<a href="/tags/主题个性化/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;主题个性化</a>
<a href="/tags/Hexo/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;Hexo</a>
<a href="/tags/Material-X/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;Material X</a>
<a href="/tags/spfk/" rel="nofollow"><i class="fas fa-tags fa-fw"></i>&nbsp;spfk</a>
......
Markdown is supported
0% 或 .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑！
想要评论请 注册 或