醫(yī)學(xué)統(tǒng)計(jì)學(xué)與SAS軟.ppt

SAS軟件應(yīng)用,Sas數(shù)據(jù)分析過(guò)程,建立數(shù)據(jù)集 整理數(shù)據(jù)集 分析數(shù)據(jù)、按照適當(dāng)?shù)男问捷敵?一、sas數(shù)據(jù)集與數(shù)據(jù)文件,1、數(shù)據(jù)集的建立 數(shù)據(jù)步 data a1; 數(shù)據(jù)步開(kāi)始 input x y; 輸入的變量 cards; 數(shù)據(jù)塊開(kāi)始（也可用infile 1 2 指定數(shù)據(jù)文件存儲(chǔ)位置） 3 4 ; 數(shù)據(jù)塊結(jié)束，必須單獨(dú)占一行 run; 數(shù)據(jù)步結(jié)束語(yǔ)句（可以省略） Infile e:tmp.dat;,過(guò)程步生成數(shù)據(jù)集（很多sas過(guò)程都可以） proc means data=sashelp.class noprint; var weight; output out=a2; run; 也可以采用sas的ods生成數(shù)據(jù)集。,2、sas與其他數(shù)據(jù)文件格式的導(dǎo)入和導(dǎo)出 導(dǎo)入數(shù)據(jù) import過(guò)程 proc import out=a3 datafile=e:tmp.xls dbms=excel replace; run; 導(dǎo)出數(shù)據(jù)export過(guò)程 proc export data=sashelp.class outfile=e:stu.csv dbms=csv replace; run;,二、數(shù)據(jù)集整理(數(shù)據(jù)加工),1、變量類型和缺失值 數(shù)值型 缺失值用 ” . ” 表示 字符型 缺失值用 ”空格“ 表示 日期型 自定義型 根據(jù)自定義的類型屬于字符還是數(shù)值,data a4; infile e:stu.csv delimiter = , dsd firstobs=2; input name:$20. sex:$6. age height weight; bmi=(weight/2)/(height/100); 數(shù)據(jù)加工語(yǔ)句 run;,data a5; set sashelp.clss; bmi=(weight/2)/(height/100);數(shù)據(jù)加工語(yǔ)句 run;,2、數(shù)據(jù)集選項(xiàng),data a6; set sashelp.class(keep=weight); run; 以上程序的作用是生成的數(shù)據(jù)集中僅包含sashelp.class中得weight變量，常用的選項(xiàng)有keep drop rename where等,3、對(duì)觀測(cè)的選擇 If語(yǔ)句 If 表達(dá)式 then 語(yǔ)句; else 語(yǔ)句;,data a6; set sashelp.class(keep=weight); if weight100 then w=低體重; else w=高體重; run; 邏輯表達(dá)式為1時(shí)執(zhí)行then后邊的語(yǔ)句，否則不執(zhí)行后邊的語(yǔ)句。 上例weight100為時(shí)執(zhí)行 w=低體重 其他的執(zhí)行 w=高體重,4、常用的運(yùn)算符,data a6; set sashelp.class; if (sex=男) then w=低體重; else w=高體重; run;,data a6; set sashelp.class; if sex=男 and age14 then m=1; run;,5、where 表達(dá)式 data a7; set sashelp.class(where=(sex=男); run; proc print data=a7; run;,6、按變量值對(duì)數(shù)據(jù)集排序,proc sort data=sashelp.class out=a7; by age; by descending weight; run;,7、用sas語(yǔ)句生成新變量 a 賦值語(yǔ)句 b 使用sas函數(shù) Total=sum（weight，height）,data a4; infile e:stu.csv delimiter = , dsd firstobs=2; input name:$20. sex:$6. age height weight; bmi=(weight/2)/(height/100);（賦值語(yǔ)句） run;,c 求和語(yǔ)句 data a8 ; set sashelp.class; i+1; run;,8、do end語(yǔ)句,data a8 ; set sashelp.class; if sex=男 then do ; a=1; b=1; end; run;,9、sas函數(shù),常用的函數(shù),data a4; infile e:stu.csv delimiter = , dsd firstobs=2; input name:$20. sex:$6. age height weight; bmi=(weight/2)/(height/100); bmi2=round(bmi,0.01) ; run;,data a9; name = tomas; name2 = upcase(name); run; proc print data=a9; run;,data a9; x = good bad middle high; x1 = scan(x,1, ); x2 = scan(x,2); x3 = scan(x,3); x4 = scan(x,4); run;,data a9; x1 = good; x2 = bad; x3 = middle; x4 = x1 | | x2 | | x3; run; proc print data=a9; run;,10、循環(huán)語(yǔ)句,data a10 ; do a=1 to 20 ; x = a*10; output; end; run;,data a10 ; do a=1 to 20 by 2; x = a*10; output; end; run;,data a10 ; do a=1 to 20 by 2 unt