Bioinformatics home

日志

关于我

xiaofeng1982

文章分类

LOFTER精选

八招诀窍，教你实力撩妹 >

网易考拉推荐

SPSS canonical analysis

2012-11-16 05:58:37| 分类：生物统计 | 标签： |举报 |字号大中小订阅

下载LOFTER 我的照片书 |

1. 按File→New→Syntax的顺序新建一个语句窗口。在语句窗口中输入下面的语句：

INCLUDE 'C:\Program Files\IBM\SPSS\Statistics\20\Samples\English\Canonical correlation.sps'.

CANCORR SET1=x1 x2 x3 /

SET2= x4 x5 x6 / .

2. click Run => all

结果：

Correlations for Set-1

x1 x2 x3

x1 1.0000 .8362 .8651

x2 .8362 1.0000 .7863

x3 .8651 .7863 1.0000

Correlations for Set-2

x4 x5 x6

x4 1.0000 .7403 .8086

x5 .7403 1.0000 .8538

x6 .8086 .8538 1.0000

Correlations Between Set-1 and Set-2

x4 x5 x6

x1 .8089 .8800 .8407

x2 .8897 .8213 .8458

x3 .7685 .7786 .7200

Canonical Correlations

1 .954

2 .346

3 .212

上面是提取出的两个典型相关系数的大小，可见第一典型相关系数为0.954，第二典型相关系数为0.346,第三典型相关系数为0.212

Test that remaining correlations are zero:

Wilk's Chi-SQ DF Sig.

1 .075 187.428 9.000 .000

2 .841 12.586 4.000 .013

3 .955 3.340 1.000 .068

上表为检验各典型相关系数有无统计学意义，可见第一典型相关系数有统计学意义（p-value: .000) ，而第二典型相关系数 (p-value=.013)。

Standardized Canonical Coefficients for Set-1

1 2 3

x1 -.460 2.065 -.920

x2 -.563 -1.563 -.865

x3 -.023 -.525 1.979

上面为各典型变量与变量组1中各变量间标准化的系数列表，由此我们可以写出典型变量的转换公式(标化的)为：

第一典型变量: v1=-0.460 * X1 -0.563 * x2 -0.023 * X3

第一典型变量 v2= 2.065 * X1 -1.563 * x2 -0.525 * X3

Raw Canonical Coefficients for Set-1

1 2 3

x1 -.027 .121 -.054

x2 -.401 -1.114 -.616

x3 -.006 -.147 .552

上面为各典型变量与变量组1中各变量间未标准化的系数列表

Standardized Canonical Coefficients for Set-2

1 2 3

x4 -.482 -1.459 .778

x5 -.430 1.269 1.412

x6 -.165 .201 -2.209

上面为各典型变量与变量组2中各变量间标准化的系数列表，由此我们可以写出典型变量的转换公式(标化的)为：

第一典型变量: v1= -0.482 * X4 - 0.430 * x5 - 0.165 * X6

第一典型变量 v2= -1.459 * X4 + 1.269 * x5 +0.201 * X6

Raw Canonical Coefficients for Set-2

1 2 3

x4 -.143 -.432 .230

x5 -.173 .510 .567

x6 -.161 .196 -2.158

上面为各典型变量与变量组2中各变量间未标准化的系数列表

Canonical Loadings for Set-1

1 2 3

x1 -.950 .304 .070

x2 -.965 -.249 -.077

x3 -.863 .032 .504

Cross Loadings for Set-1

1 2 3

x1 -.907 .105 .015

x2 -.921 -.086 -.016

x3 -.824 .011 .107

Canonical Loadings for Set-2

1 2 3

x4 -.933 -.358 .037

x5 -.927 .360 .102

x6 -.921 .104 -.374

Cross Loadings for Set-2

1 2 3

x4 -.890 -.124 .008

x5 -.885 .125 .022

x6 -.879 .036 -.079

Redundancy Analysis:

冗余度(Redundancy)分析结果，它列出各典型相关系数所能解释原变量变异的比例，可以用来辅助判断需要保留多少个典型相关系数。

Proportion of Variance of Set-1 Explained by Its Own Can. Var.

Prop Var

CV1-1 .860

CV1-2 .052

CV1-3 .088

首先输出的是第一组变量的变异可被自身的典型变量所解释的比例，可见第一典型变量解释了总变异的86.0％，而第二典型变量只能解释5.20％。

Proportion of Variance of Set-1 Explained by Opposite Can.Var.

Prop Var

CV2-1 .783

CV2-2 .006

CV2-3 .004

第一组变量的变异能被它们相对的典型变量（即第二组的典型变量）所解释的比例，例，对于第一典型变量来说第一组变量可以被第二组变量的第一典型变量解释78.3%。

Proportion of Variance of Set-2 Explained by Its Own Can. Var.

Prop Var

CV2-1 .860

CV2-2 .090

CV2-3 .051

第二组变量的变异可被自身的典型变量所解释的比例，可见第一典型变量解释了总变异的86.0％，而第二典型变量只能解释9.0％。

Proportion of Variance of Set-2 Explained by Opposite Can. Var.

Prop Var

CV1-1 .783

CV1-2 .011

CV1-3 .002

第二组变量的变异能被它们相对的典型变量（即第一组的典型变量）所解释的比例，例，对于第一典型变量来说第二组变量可以被第一组变量的第一典型变量解释78.3%。

评论这张

转发至微博

阅读(1465)| 评论(0)

推荐转载

历史上的今天

在LOFTER的更多文章

this.p={  m:2,
              b:2,
              loftPermalink:'',
              id:'fks_087065084081084066081086080064072085080070083084082070092',
              blogTitle:'SPSS canonical analysis',
              blogAbstract:'<div\>1. 按File→New→Syntax的顺序新建一个语句窗口。在语句窗口中输入下面的语句：</div\><div\>INCLUDE \'C:\\Program Files\\IBM\\SPSS\\Statistics\\20\\Samples\\English\\Canonical correlation.sps\'.</div\><div\><span style=\"white-space:pre;\"  \> </span\>CANCORR SET1=x1 x2 x3 /</div\><div\><span style=\"white-space:pre;\"  \> </span\>SET2= x4 x5 x6 / . </div\><div\><br\></div\><div\>2. click Run => all</div\><div\><br\></div\><div\>结果：</div\><div\>Correlations for Set-1</div\>',
              blogTag:'',
              blogUrl:'blog/static/315724582012101655837400',
              isPublished:1,
              istop:false,
              type:2,
              modifyTime:1353017093482,
              publishTime:1353016717400,
              permalink:'blog/static/315724582012101655837400',
              commentCount:0,
              mainCommentCount:0,
              recommendCount:0,
              bsrk:-100,
              publisherId:0,
              recomBlogHome:false,
              currentRecomBlog:false,
              attachmentsFileIds:[],
              vote:{},
              groupInfo:{},
              friendstatus:'none',
              followstatus:'unFollow',
              pubSucc:'',
              visitorProvince:'',
              visitorCity:'',
              visitorNewUser:false,
              postAddInfo:{},
              mset:'000',
              mcon:'',
              srk:-100,
              remindgoodnightblog:false,
              isBlackVisitor:false,
              isShowYodaoAd:true,
              hostIntro:'',
              hmcon:'1',
              selfRecomBlogCount:'0',
              lofter_single:'<a hidefocus="true" target="_blank" href="http://inpin.blog.163.com/blog/static/2633040402016620112520219/"><img onload="var nimg = new Image(); nimg.src=\'http://blog.163.com/newpage/images/analyse.png?lofter_single&t=\'+new Date().getTime()" border="0" src="http://imglf.nosdn.127.net/img/TlZCWHdFWFpidkQ1dGtJK1N5cFRBNzBLUFFPYnJEVi9LWk1KTzJlMTU2Yz0.jpg"/></a>'
            }

{list a as x}
    {if !!x}
    <div class="iblock nbw-fce nbw-f40">
      <a class="fc03 noul" target="_blank" hidefocus="true" href="http://blog.163.com/${x.visitorName}/">
      {if x.visitorName==visitor.userName}
      <img alt="${x.visitorNickname|escape}" onerror="this.src=location.f40" class="cwd bdwa bdc0" src="${fn1(x.visitorName)}&r=${visitor.imageUpdateTime}"/>
      {else}
      <img alt="${x.visitorNickname|escape}" onerror="this.src=location.f40" class="cwd bdwa bdc0" src="${fn1(x.visitorName)}"/>
      {/if}
      </a>
      <div class="cwd vname thide">
        {if x.moveFrom=='wap'}
          <a class="noul pnt" target="_blank" href="http://blog.163.com/services/wapblog.html?frompersonalbloghome"><span title="来自网易手机博客" class="iblock wapIcon"> </span></a>
        {elseif x.moveFrom=='iphone'}
          <a class="noul pnt" target="_blank"><span title="来自iPhone客户端" class="iblock iphoneIcon"> </span></a>
        {elseif x.moveFrom=='android'}
          <a class="noul pnt" target="_blank"><span title="来自Android客户端" class="iblock androidIcon"> </span></a>
        {elseif x.moveFrom=='mobile'}
          <a class="noul pnt" target="_blank" href="http://blog.163.com/services/emsblog.html?frompersonalbloghome"><span title="来自网易短信写博" class="iblock wapIcon"> </span></a>
        {/if}
        <a class="fc03 m2a"  target="_blank" hidefocus="true" href="http://blog.163.com/${x.visitorName}/">
          ${fn(x.visitorNickname,8)|escape}
        </a>
      </div>
    </div>
    {/if}
    {/list}

<#--最新日志，群博日志--> <#--推荐日志-->

<p class="fc06">推荐过这篇日志的人：</p>
    <div>
      {list a as x}
      {if !!x}
      <div class="iblock nbw-fce nbw-f40">
        <a class="fc03 noul" target="_blank" hidefocus="true" href="http://blog.163.com/${x.recommenderName}/">
        <img alt="${x.recommenderNickname|escape}" onerror="this.src=location.f40" class="cwd bdwa bdc0" src="${fn1(x.recommenderName)}"/>
        </a>
        <div class="cwd thide">
          <a class="fc03 m2a" target="_blank" hidefocus="true" href="http://blog.163.com/${x.recommenderName}/">
            ${fn(x.recommenderNickname,6)|escape}
          </a>
        </div>
      </div>
      {/if}
      {/list}
    </div>
    {if !!b&&b.length>0}
    <p  class="fc06">他们还推荐了：</p>
    <ul>
    {list b as y}
      {if !!y}
        <li class="rrb"><span class="iblock">·</span><a class="fc03 m2a" target="_blank" href="http://blog.163.com/${y.recommendBlogPermalink}/?from=blog/static/315724582012101655837400">${y.recommendBlogTitle|escape}</a></li>
      {/if}
    {/list}
    </ul>
    {/if}

<#--引用记录--> <#--博主推荐--> <#--随机阅读--> <#--首页推荐--> <#--历史上的今天--> <#--被推荐日志--> <#--上一篇，下一篇--> <#-- 热度 -->

{list a as x}
    {if !!x}
    <div class="hotItem iblock nbw-fce nbw-f40">
      <a class="fc03 noul" target="_blank" hidefocus="true" href="http://blog.163.com/${x.publisherUsername}/">
      {if x.publisherUsername==visitor.userName}
      <img alt="${x.publisherNickname|escape}" onerror="this.src=location.f40" class="cwd bdwa bdc0" src="${fn1(x.publisherUsername)}&r=${visitor.imageUpdateTime}"/>
      {else}
      <img alt="${x.publisherNickname|escape}" onerror="this.src=location.f40" class="cwd bdwa bdc0" src="${fn1(x.publisherUsername)}"/>
      {/if}
      </a>
      <div class="cwd vname thide">
        <a class="fc03 m2a"  target="_blank" hidefocus="true" href="http://blog.163.com/${x.publisherUsername}/">
          ${fn(x.publisherNickname,8)|escape}
        </a>
      </div>
      <a class="f-myLikeIcons hottype {if x.type==1} js-liketype{elseif x.type==2} js-reblogtype{elseif x.type==3} js-sharetype{else}{/if}" target="_blank" hidefocus="true" href="http://blog.163.com/${x.publisherUsername}/"> </a>
    </div>
    {/if}
    {/list}

<#-- 网易新闻广告 --> <#--右边模块结构--> <#--评论模块结构--> <#--引用模块结构--> <#--博主发起的投票-->

页脚

我的照片书 - 博客风格 - 手机博客 - 下载LOFTER APP - 订阅此博客

Bioinformatics home

导航

日志

SPSS canonical analysis

历史上的今天

最近读者

热度

在LOFTER的更多文章

评论

页脚