SAS进阶《深入分析SAS》之数据汇总和展现

SAS进阶《深入分析SAS》之数据汇总和展现

1. 通过Print过程制作报表

proc print <data=数据集>;
run;
选项：
obs=修改观测序号列标签
noobs不显示观测序列号
id语句在输出中取代观测序列
var选择输出的变量
where语句选择输出的观测

总结如下：
    proc print data=数据集<选项>;
        id 变量1<变量2...>;
        var 变量1<变量2...>;
        where 表达式;
        sum 变量1 <变量2...>;
        where 表达式;
        sum 变量1<变量2...>;
        sumby 变量1<变量2...>;
    run;

2. tabulate

proc tabulate data=Saslib.Sales2;
    title1 'Sales in North America';
    title2 'TOtal Transactions';
/*
*class为分类变量
*/
    class Emp_ID Dept;
    var Sales;
    table Dept*Emp_ID,Sales;
run;

3. 通过GPLOT过程制作图行

分组变量

    axis1 order=(18900 to 18960 by 5);
    axis2 order=(9000 to 20000 by 1000);
        minor=(color=blue height=0.25 number=1); 
    symbol value=# cv=red
            interpol=join ci=blue;
    proc gplot data = Saslib.Sales;
        title f = 'Albany Amt' c = blue h = 3 u = 2 'Yearly Amount in North America';
        footnote j = r 'Optimization Solution Co.Ltd';
        plot Sales*Date=Emp_ID/haxis=axis1 vaxis=axis2;
    run;
    quit;
    goptions reset=all;

4. 增强型HTLM输出

    proc sort data=sashelp.prdsale out=work.prdsale;
        by country;
    run;
    ods html path="C:UsersLEIDesktopdata"
             body='prdsalebody.html'
             frame='prdsaleframe.html'
             contents='prdsalecontents.html';

    proc tabulate data=work.prdsale;
        class region division prodtype;
        var actual;keyword all sum;
        keylabel all='Total';
        table (region all)*(division all),
              (prodtype all)*(actual*f=dollar10.) / misstext=[label='Missing']
              box=[label='Region by Division and Type'];
    run;

    ods select ExtremeObs Quantiles Moments;
    proc univariate data=work.prdsale;
        by Country;
        var actual;
    run;

    ods html close;

总结：SAS对数据的汇总和表现两种方式：一种是列表，一种是图行。通过print和tabulate过程制作各种报表和汇总报表；GPLOT过程和GCHART过制作散点图、连线图、气泡图、柱状图等多种图行；ODS输出传送系统，包括如何选择或挑剔输出对象，创建多种格式的输出文件。