<html xmlns="http://www.w3.org/1999/xhtml">
<head>
<meta charset="utf-8">
<meta name="generator" content="pdf2htmlEX">
<meta http-equiv="X-UA-Compatible" content="IE=edge,chrome=1">
<link rel="stylesheet" href="https://static.pudn.com/base/css/base.min.css">
<link rel="stylesheet" href="https://static.pudn.com/base/css/fancy.min.css">
<link rel="stylesheet" href="https://static.pudn.com/prod/directory_preview_static/626b80da7ae5df2aa71475a0/raw.css">
<script src="https://static.pudn.com/base/js/compatibility.min.js"></script>
<script src="https://static.pudn.com/base/js/pdf2htmlEX.min.js"></script>
<script>
try{
pdf2htmlEX.defaultViewer = new pdf2htmlEX.Viewer({});
}catch(e){}
</script>
<title></title>
</head>
<body>
<div id="sidebar" style="display: none">
<div id="outline">
</div>
</div>
<div id="pf1" class="pf w0 h0" data-page-no="1"><div class="pc pc1 w0 h0"><img class="bi x0 y0 w1 h1" alt="" src="https://static.pudn.com/prod/directory_preview_static/626b80da7ae5df2aa71475a0/bg1.jpg"><div class="t m0 x1 h2 y1 ff1 fs0 fc0 sc0 ls0 ws0"> </div><div class="t m0 x2 h2 y2 ff1 fs0 fc0 sc0 ls0 ws0"> </div><div class="t m0 x2 h3 y3 ff2 fs1 fc0 sc1 ls0 ws0">数据描述性分析<span class="ff3 sc0"> </span></div><div class="t m0 x3 h4 y4 ff4 fs1 fc0 sc0 ls0 ws0"><span class="ff5"> <span class="_ _0"> </span><span class="ff2">一个数据很可能有很多变量和很多观测值。<span class="_ _1"></span>但那么多的数字也可以用一些简<span class="_ _1"></span>单的表</span></span></div><div class="t m0 x4 h5 y5 ff2 fs1 fc0 sc0 ls0 ws0">格、<span class="_ _2"></span>图形<span class="_ _2"></span>和少数<span class="_ _2"></span>汇总<span class="_ _2"></span>数字来<span class="_ _2"></span>描述。<span class="_ _2"></span>这些<span class="_ _2"></span>描述的<span class="_ _2"></span>方法被<span class="_ _2"></span>称为描<span class="_ _2"></span>述统<span class="_ _2"></span>计学（<span class="_ _3"></span><span class="ff1">descri<span class="_ _1"></span>ptive </span></div><div class="t m0 x4 h5 y6 ff1 fs1 fc0 sc0 ls0 ws0">statistics<span class="_ _1"></span>)<span class="ff2">，也可以称为探<span class="_ _1"></span>索性数据分析<span class="ff1">(E<span class="_ _1"></span>DA,explora<span class="_ _1"></span>tory data analy<span class="_ _1"></span>sis)<span class="ff2">。</span> </span></span></div><div class="t m0 x3 h4 y7 ff4 fs1 fc0 sc0 ls0 ws0"><span class="ff5"> <span class="_ _0"> </span><span class="ff2">描述统计的目的在于帮助展示和理解数据，<span class="_ _1"></span>而且对数据的特征进行探索、在<span class="_ _1"></span>报刊、</span></span></div><div class="t m0 x4 h5 y8 ff2 fs1 fc0 sc0 ls0 ws0">杂志、网络等各<span class="_ _1"></span>媒体中所出现的<span class="_ _1"></span>表格、数字和<span class="_ _1"></span>图形大多都是对<span class="_ _1"></span>数据的描述。<span class="_ _1"></span><span class="ff1"> </span></div><div class="t m0 x3 h4 y9 ff4 fs1 fc0 sc0 ls0 ws0"><span class="ff5"> <span class="_ _0"> </span><span class="ff2">数据探索的主要作用是利用人的直觉来识别<span class="_ _1"></span>模式，实际上人们往往能够识别<span class="_ _1"></span>数据分</span></span></div><div class="t m0 x4 h5 ya ff2 fs1 fc0 sc0 ls0 ws0">析工具所不能发<span class="_ _1"></span>现的模式，此外<span class="_ _1"></span>它能帮助选择<span class="_ _1"></span>适当的处理和分<span class="_ _1"></span>析方法。<span class="ff1"> </span></div><div class="t m0 x3 h4 yb ff4 fs1 fc0 sc0 ls0 ws0"><span class="ff5"> <span class="_ _0"> </span><span class="ff2">已经一组试验<span class="_ _4"></span>（或观测）<span class="_ _4"></span>数据为：<span class="_ _5"> </span>它们可以是从所<span class="_ _1"></span>要研究的对</span></span></div><div class="t m0 x4 h5 yc ff2 fs1 fc0 sc0 ls0 ws0">象的全体<span class="ff1">-<span class="_ _1"></span><span class="ff2">总体<span class="_ _6"> </span><span class="ff1">X<span class="_"> </span></span>中取出的<span class="_ _1"></span>，<span class="_ _7"></span>这<span class="_ _6"> </span><span class="ff1">n<span class="_"> </span></span>个观测值<span class="_ _1"></span>就构成一个样本<span class="_ _1"></span>。<span class="_ _7"></span>在某些<span class="_ _1"></span>简单的实际问题</span></span></div><div class="t m0 x4 h5 yd ff2 fs1 fc0 sc0 ls0 ws0">中，这<span class="_ _6"> </span><span class="ff1">n<span class="_ _8"> </span></span>个观测值就<span class="_ _1"></span>是所要研究问题<span class="_ _1"></span>的全体。数据<span class="_ _1"></span>分析的任务就是<span class="_ _1"></span>要对这全部<span class="_ _6"> </span><span class="ff1">n<span class="_ _8"> </span></span>个</div><div class="t m0 x4 h5 ye ff2 fs1 fc0 sc0 ls0 ws0">数据进行分析，<span class="_ _1"></span>提取数据中包含<span class="_ _1"></span>的有用信息。<span class="_ _1"></span><span class="ff1"> </span></div><div class="t m0 x3 h4 yf ff4 fs1 fc0 sc0 ls0 ws0"><span class="ff5"> <span class="_ _0"> </span><span class="ff2">数据作为信息的载体，当然要分析数据中包<span class="_ _1"></span>含的主要信息，即要分析数据的<span class="_ _1"></span>主要特</span></span></div><div class="t m0 x4 h6 y10 ff2 fs1 fc0 sc0 ls0 ws0">征。也就是说，要研究数据的数字特征。对<span class="_ _1"></span>于数据的数字特征，要分析数据<span class="_ _1"></span>的集中</div><div class="t m0 x4 h5 y11 ff2 fs1 fc0 sc0 ls0 ws0">位置、分散程度<span class="_ _1"></span>和数据分布等。<span class="_ _1"></span><span class="ff1"> </span></div><div class="t m0 x2 h3 y12 ff3 fs1 fc0 sc0 ls0 ws0"> <span class="_ _9"> </span> </div><div class="t m0 x2 h3 y13 ff2 fs1 fc0 sc1 ls0 ws0">位置的度量<span class="ff3 sc0"> </span></div><div class="t m0 x3 h4 y14 ff4 fs1 fc0 sc0 ls0 ws0"><span class="ff5"> <span class="_ _0"> </span><span class="ff2">有些汇总统计量是描述数据“位置”的。其<span class="_ _1"></span>实数据的每个点都有自己的位置<span class="_ _1"></span>，不可</span></span></div><div class="t m0 x4 h6 y15 ff2 fs1 fc0 sc0 ls0 ws0">能都一一列举；能够做的就是描述数据的“<span class="_ _1"></span>中间”或“中心”在哪里，或者<span class="_ _1"></span>数据的</div><div class="t m0 x4 h5 y16 ff2 fs1 fc0 sc0 ls0 ws0">百分之几的数据<span class="_ _1"></span>点小于哪个数等<span class="_ _1"></span>等。<span class="ff1"> <span class="_ _8"> </span> </span></div><div class="t m0 x3 h4 y17 ff4 fs1 fc0 sc0 ls0 ws0"><span class="ff5"> <span class="_ _0"> </span><span class="ff2">所谓位置的度量<span class="_ _1"></span>就是用来描述定<span class="_ _1"></span>量资料的集中<span class="_ _1"></span>趋势的统计量<span class="_ _1"></span>，<span class="_ _a"></span>常用的有均<span class="_ _1"></span>值、<span class="_ _a"></span>众数、</span></span></div><div class="t m0 x4 h5 y18 ff2 fs1 fc0 sc0 ls0 ws0">中位数、百分位<span class="_ _1"></span>数等。这些指标<span class="_ _1"></span>的主要作用包<span class="_ _1"></span>括：<span class="ff1"> </span></div><div class="t m0 x3 h5 y19 ff6 fs1 fc0 sc0 ls0 ws0">•<span class="ff5"> <span class="_ _b"> </span><span class="ff2">反映总体各单位<span class="_ _1"></span>变量分布的集中<span class="_ _1"></span>趋势和一般水<span class="_ _1"></span>平；<span class="ff1"> </span></span></span></div><div class="t m0 x3 h5 y1a ff6 fs1 fc0 sc0 ls0 ws0">•<span class="ff5"> <span class="_ _b"> </span><span class="ff2">便于比较同类现<span class="_ _1"></span>象在不同单位之<span class="_ _1"></span>间的水平；<span class="_ _1"></span><span class="ff1"> </span></span></span></div><div class="t m0 x3 h5 y1b ff6 fs1 fc0 sc0 ls0 ws0">•<span class="ff5"> <span class="_ _b"> </span><span class="ff2">便于比较同类现<span class="_ _1"></span>象在不同时期的<span class="_ _1"></span>发展变化趋势<span class="_ _1"></span>或规律；<span class="ff1"> </span></span></span></div><div class="t m0 x3 h5 y1c ff6 fs1 fc0 sc0 ls0 ws0">•<span class="ff5"> <span class="_ _b"> </span><span class="ff2">用于分析现象之<span class="_ _1"></span>间的依存关系。<span class="_ _1"></span><span class="ff1"> </span></span></span></div><div class="t m0 x2 h3 y1d ff2 fs1 fc0 sc1 ls0 ws0">均值<span class="ff3 sc0"> </span></div><div class="t m0 x5 h3 y1e ff3 fs1 fc0 sc0 ls0 ws0"> </div><div class="t m0 x2 h3 y1f ff2 fs1 fc0 sc1 ls0 ws0">例：<span class="ff3 sc0">mean( )</span>函数应用<span class="_ _1"></span><span class="ff3 sc0"> </span></div><div class="t m0 x3 h5 y20 ff4 fs1 fc0 sc0 ls0 ws0"><span class="ff5"> <span class="_ _0"> </span><span class="ff2">利用<span class="_ _6"> </span><span class="ff1">mean</span>（<span class="_ _1"></span>）函数求解均<span class="_ _1"></span>值：<span class="ff1"> </span></span></span></div><div class="t m0 x6 h5 y21 ff1 fs1 fc0 sc0 ls0 ws0"> </div></div><div class="pi" data-data='{"ctm":[1.611792,0.000000,0.000000,1.611792,0.000000,0.000000]}'></div></div>
</body>
</html>
<div id="pf2" class="pf w0 h0" data-page-no="2"><div class="pc pc2 w0 h0"><img class="bi x0 y0 w1 h1" alt="" src="https://static.pudn.com/prod/directory_preview_static/626b80da7ae5df2aa71475a0/bg2.jpg"><div class="t m0 x1 h2 y1 ff1 fs0 fc0 sc0 ls0 ws0"> </div><div class="t m0 x2 h2 y2 ff1 fs0 fc0 sc0 ls0 ws0"> </div><div class="t m0 x3 h5 y3 ff4 fs1 fc0 sc0 ls0 ws0"><span class="ff5"> <span class="_ _0"> </span><span class="ff2">当<span class="_ _6"> </span><span class="ff1">x<span class="_"> </span></span>是矩阵<span class="_ _1"></span>（<span class="_ _1"></span>或数组）<span class="_ _1"></span>时<span class="_ _1"></span>，<span class="_ _1"></span>函数<span class="_ _6"> </span><span class="ff1">m<span class="_ _1"></span>ean<span class="_ _1"></span><span class="ff2">（）<span class="_ _1"></span>的返回<span class="_ _1"></span>值，<span class="_ _1"></span>并不是<span class="_ _1"></span>向量，<span class="_ _1"></span>而是一<span class="_ _1"></span>个数，<span class="_ _4"></span>即</span></span></span></span></div><div class="t m0 x4 h5 y4 ff2 fs1 fc0 sc0 ls0 ws0">矩阵中全部数据<span class="_ _1"></span>的平均值。例如<span class="_ _1"></span>：<span class="ff1"> </span></div><div class="t m0 x7 h5 y22 ff1 fs1 fc0 sc0 ls0 ws0"> </div><div class="t m0 x3 h5 y23 ff4 fs1 fc0 sc0 ls0 ws0"><span class="ff5"> <span class="_ _0"> </span><span class="ff2">如果需要得到矩<span class="_ _1"></span>阵各行或各列的<span class="_ _1"></span>均值，需要调<span class="_ _1"></span>用<span class="_ _6"> </span><span class="ff1">apply<span class="_ _1"></span><span class="ff2">（）函数。<span class="ff1"> </span></span></span></span></span></div><div class="t m0 x8 h5 y24 ff1 fs1 fc0 sc0 ls0 ws0"> </div><div class="t m0 x2 h5 y11 ff2 fs1 fc0 sc0 ls0 ws0">设置<span class="_ _6"> </span><span class="ff1">trim<span class="_ _c"> </span></span>参数<span class="ff1"> </span></div><div class="t m0 x3 h5 y12 ff4 fs1 fc0 sc0 ls0 ws0"><span class="ff5"> <span class="_ _0"> </span><span class="ff2">如果研究的数据<span class="_ _1"></span>中存在异常值，<span class="_ _1"></span>这时候就不能<span class="_ _1"></span>简单用<span class="_ _d"> </span><span class="ff1">mean</span>（<span class="ff1">x</span>）<span class="_ _1"></span>计算平均值<span class="_ _1"></span>了。可</span></span></div><div class="t m0 x4 h5 y13 ff2 fs1 fc0 sc0 ls0 ws0">以通过设置<span class="_ _c"> </span><span class="ff1">trim<span class="_ _e"> </span></span>参数来调整纳入<span class="_ _1"></span>计算的样本数<span class="_ _1"></span>据来实现剔除异<span class="_ _1"></span>常值后再求平<span class="_ _1"></span>均的效</div><div class="t m0 x4 h5 y14 ff2 fs1 fc0 sc0 ls0 ws0">果。<span class="ff1"> </span></div><div class="t m0 x3 h5 y15 ff4 fs1 fc0 sc0 ls0 ws0"><span class="ff5"> <span class="_ _0"> </span><span class="ff2">依旧以上面的数<span class="_ _1"></span>据<span class="_ _6"> </span><span class="ff1">w<span class="_"> </span></span>为研<span class="_ _1"></span>究对象，<span class="_ _1"></span>如果<span class="_ _c"> </span><span class="ff1">w<span class="_"> </span></span>的第一个<span class="_ _1"></span>元素改为<span class="_ _6"> </span><span class="ff1">750</span>，<span class="_ _4"></span>则是个明显的异</span></span></div><div class="t m0 x4 h5 y16 ff2 fs1 fc0 sc0 ls0 ws0">常值。这时候我<span class="_ _1"></span>们可以通过设<span class="_ _1"></span>置<span class="_ _6"> </span><span class="ff1">trim<span class="_ _c"> </span></span>参数的值来实现剔除<span class="_ _1"></span>异常元素后再<span class="_ _1"></span>求均值。<span class="ff1"> </span></div><div class="t m0 x9 h5 y25 ff1 fs1 fc0 sc0 ls0 ws0"> </div><div class="t m0 x2 h5 y1b ff2 fs1 fc0 sc0 ls0 ws0">其中<span class="_ _6"> </span><span class="ff1">trim<span class="_ _c"> </span></span>的取值在<span class="_ _c"> </span><span class="ff1">0<span class="_"> </span></span>值<span class="_ _6"> </span><span class="ff1">0.5<span class="_"> </span></span>之间，表示在<span class="_ _1"></span>计算均值前需要<span class="_ _1"></span>去掉异常值的<span class="_ _1"></span>比例。<span class="ff1"> </span></div><div class="t m0 x2 h5 y1c ff2 fs1 fc0 sc0 ls0 ws0">利用这个参数可<span class="_ _1"></span>以有效的改善异<span class="_ _1"></span>常值对计算的<span class="_ _1"></span>影响。<span class="ff1"> </span></div><div class="t m0 x2 h5 y1d ff2 fs1 fc0 sc0 ls0 ws0">注意：<span class="ff1">trim<span class="_ _c"> </span></span>参数是对排<span class="_ _1"></span>序后的数据从<span class="_ _1"></span>头尾剔除相同<span class="_ _1"></span>个数的元素后再<span class="_ _1"></span>求均值的。<span class="ff1"> </span></div><div class="t m0 x2 h5 y26 ff2 fs1 fc0 sc0 ls0 ws0">验证：<span class="ff1">0.1*<span class="_ _1"></span>15=1.5<span class="ff2">，即说明<span class="_ _1"></span>一共要剔除两个<span class="_ _1"></span>数据，分别是样<span class="_ _1"></span>本数据的最小<span class="_ _1"></span>值和最大值<span class="ff1"> </span></span></span></div><div class="t m0 xa h5 y27 ff1 fs1 fc0 sc0 ls0 ws0"> </div><div class="t m0 x2 h5 y20 ff2 fs1 fc0 sc0 ls0 ws0">设置<span class="_ _6"> </span><span class="ff1">na.rm<span class="_ _c"> </span></span>参数<span class="ff1"> </span></div><div class="t m0 x3 h5 y28 ff4 fs1 fc0 sc0 ls0 ws0"><span class="ff5"> <span class="_ _0"> </span><span class="ff1">na.rm<span class="_"> </span><span class="ff2">是控制缺<span class="_ _1"></span>失数据的参<span class="_ _1"></span>数。<span class="ff1"> </span></span></span></span></div><div class="t m0 x3 h5 y29 ff4 fs1 fc0 sc0 ls0 ws0"><span class="ff5"> <span class="_ _0"> </span><span class="ff2">例如一个向量<span class="_ _1"></span>有<span class="_ _8"> </span><span class="ff1 ls1">16<span class="_ _8"> </span></span>个<span class="_ _1"></span>元素，但是第<span class="_ _6"> </span><span class="ff1">16<span class="_ _8"> </span></span>个元<span class="_ _1"></span>素缺失，如果按<span class="_ _1"></span>照通常的计算<span class="_ _1"></span>方法，将</span></span></div><div class="t m0 x4 h5 y2a ff2 fs1 fc0 sc0 ls0 ws0">得不到结果。<span class="_ _1"></span><span class="ff1"> </span></div></div><div class="pi" data-data='{"ctm":[1.611792,0.000000,0.000000,1.611792,0.000000,0.000000]}'></div></div>
<div id="pf3" class="pf w0 h0" data-page-no="3"><div class="pc pc3 w0 h0"><img class="bi x0 y0 w1 h1" alt="" src="https://static.pudn.com/prod/directory_preview_static/626b80da7ae5df2aa71475a0/bg3.jpg"><div class="t m0 x1 h2 y1 ff1 fs0 fc0 sc0 ls0 ws0"> </div><div class="t m0 x2 h2 y2 ff1 fs0 fc0 sc0 ls0 ws0"> </div><div class="t m0 xb h5 y2b ff1 fs1 fc0 sc0 ls0 ws0"> </div><div class="t m0 x2 h5 y6 ff4 fs1 fc0 sc0 ls0 ws0"><span class="ff5"> <span class="_ _f"> </span><span class="ff2">选用参数<span class="_ _c"> </span><span class="ff1">na.rm=TRUE<span class="_"> </span></span>可<span class="_ _1"></span>以很好的处理这<span class="_ _1"></span>个问题，看一下<span class="_ _1"></span>计算结果：<span class="_ _1"></span><span class="ff1"> </span></span></span></div><div class="t m0 xc h5 y2c ff1 fs1 fc0 sc0 ls0 ws0"> </div><div class="t m0 x2 h5 y9 ff1 fs1 fc0 sc0 ls0 ws0">weighted.m<span class="_ _1"></span>ean( )<span class="ff2">函数</span> </div><div class="t m0 x3 h5 ya ff4 fs1 fc0 sc0 ls0 ws0"><span class="ff5"> <span class="_ _0"> </span><span class="ff2">与均值函数<span class="_ _c"> </span><span class="ff1">mean<span class="_ _10"></span><span class="ff2">（）<span class="_ _10"></span>相关的函数还<span class="_ _1"></span>有<span class="_ _6"> </span><span class="ff1">weighted.m<span class="_ _1"></span>ean<span class="_ _10"></span><span class="ff2">（）<span class="_ _11"></span>。<span class="_ _10"></span>即计算数据的<span class="_ _1"></span>加权平均值。</span></span></span></span></span></span></div><div class="t m0 x4 h5 y2d ff2 fs1 fc0 sc0 ls0 ws0">具体的使用格式<span class="_ _1"></span>为：<span class="ff1"> </span></div><div class="t m0 x2 h5 y2e ff1 fs1 fc0 sc0 ls0 ws0">weighted.m<span class="_ _1"></span>ean(x, w<span class="_ _4"></span>, ..., na.rm<span class="_ _1"></span> = F<span class="_ _4"></span>ALSE) </div><div class="t m0 x2 h5 yc ff2 fs1 fc0 sc0 ls0 ws0">其中<span class="_ _6"> </span><span class="ff1">x<span class="_"> </span></span>是数值向<span class="_ _1"></span>量，<span class="ff1">w<span class="_"> </span></span>是<span class="_ _1"></span>数据的权，与<span class="_ _c"> </span><span class="ff1">x<span class="_"> </span></span>的维数相同。参<span class="_ _1"></span>数<span class="_ _6"> </span><span class="ff1">na.rm<span class="_ _c"> </span></span>的意义与<span class="ff1"> </span></div><div class="t m0 x2 h5 yd ff1 fs1 fc0 sc0 ls2 ws0"> <span class="ls0">mean<span class="ff2">（<span class="_ _1"></span>）函数相同<span class="_ _1"></span>。<span class="ff1"> </span></span></span></div><div class="t m0 x2 h5 ye ff2 fs1 fc0 sc0 ls0 ws0">该函数可以对矩<span class="_ _1"></span>阵和数组计算加<span class="_ _1"></span>权平均值，但<span class="_ _1"></span>对数据框不适用<span class="_ _1"></span>。<span class="ff1"> </span></div><div class="t m0 x5 h5 y2f ff1 fs1 fc0 sc0 ls0 ws0"> </div><div class="t m0 x2 h5 y12 ff2 fs1 fc0 sc0 ls0 ws0">实现过程：<span class="ff1"> </span></div><div class="t m0 x5 h5 y30 ff1 fs1 fc0 sc0 ls0 ws0"> </div><div class="t m0 x2 h5 y31 ff1 fs1 fc0 sc0 ls0 ws0"> </div><div class="t m0 x2 h5 y17 ff2 fs1 fc0 sc0 ls0 ws0">中位数<span class="ff1"> </span></div><div class="t m0 x3 h5 y18 ff4 fs1 fc0 sc0 ls0 ws0"><span class="ff5"> <span class="_ _0"> </span><span class="ff2">中位数（<span class="ff1">m<span class="_ _1"></span>edian<span class="ff2">，记为<span class="_ _c"> </span></span>m</span></span></span></div><div class="t m0 xd h7 y32 ff1 fs2 fc0 sc0 ls0 ws0">e</div><div class="t m0 xe h5 y18 ff2 fs1 fc0 sc0 ls0 ws0">）定义为数据排<span class="_ _1"></span>序位于中间位置<span class="_ _1"></span>的值，即<span class="ff1"> </span></div><div class="t m0 x3 h5 y33 ff4 fs1 fc0 sc0 ls0 ws0"><span class="ff5"> <span class="_ _12"> </span><span class="ff1"> </span></span></div><div class="t m0 x3 h4 y1d ff4 fs1 fc0 sc0 ls0 ws0"><span class="ff5"> <span class="_ _0"> </span><span class="ff2">中位数描述数据中心位置的数字特征。大体<span class="_ _1"></span>上比中位数大或小的数据个数为<span class="_ _1"></span>整个数</span></span></div><div class="t m0 x4 h6 y26 ff2 fs1 fc0 sc0 ls0 ws0">据的一半。对于对称分布的数据，均值与中<span class="_ _1"></span>位数比较接近；对于偏态分布的<span class="_ _1"></span>数据，</div><div class="t m0 x4 h5 y34 ff2 fs1 fc0 sc0 ls0 ws0">均值与中位数不<span class="_ _1"></span>同。<span class="ff1"> </span></div><div class="t m0 x3 h4 y35 ff4 fs1 fc0 sc0 ls0 ws0"><span class="ff5"> <span class="_ _0"> </span><span class="ff2">中位数的又一显著特征是不受异常值的影响<span class="_ _1"></span>，具有稳健性，因此它是数据分<span class="_ _1"></span>析中相</span></span></div><div class="t m0 x4 h5 y36 ff2 fs1 fc0 sc0 ls0 ws0">当重要的统计量<span class="_ _1"></span>。<span class="ff1"> </span></div><div class="t m0 x2 h5 y37 ff1 fs1 fc0 sc0 ls0 ws0">median<span class="_ _1"></span><span class="ff2">（）函数<span class="ff1"> </span></span></div><div class="t m0 x3 h5 y38 ff4 fs1 fc0 sc0 ls0 ws0"><span class="ff5"> <span class="_ _0"> </span><span class="ff2">在<span class="_ _6"> </span><span class="ff1">R<span class="_"> </span></span>软件中，<span class="_ _1"></span>函数<span class="_ _6"> </span><span class="ff1 ls3">me<span class="_ _2"></span><span class="ls0">di<span class="_ _1"></span>an<span class="ff2">（）给出观<span class="_ _1"></span>测量的中位数<span class="_ _1"></span>。如<span class="ff1"> </span></span></span></span></span></span></div><div class="t m0 xf h5 y39 ff1 fs1 fc0 sc0 ls0 ws0"> </div><div class="t m0 x10 h5 y1f ff2 fs1 fc0 sc0 ls0 ws0">实现过程：<span class="ff1"> </span></div></div><div class="pi" data-data='{"ctm":[1.611792,0.000000,0.000000,1.611792,0.000000,0.000000]}'></div></div>
