SAS基础编程.docx
《SAS基础编程.docx》由会员分享，可在线阅读，更多相关《SAS基础编程.docx（22页珍藏版）》请在冰豆网上搜索。
SAS基础编程
第三课SAS编程–第一部分
一．SAS变量的定义
/*直接赋值*/
datanewvar;
a1=100;
a2=1.2e-5;/*数值型变量的科学表达法*/
b=0100;/*数值型变量前面的0不起作用*/
c=‘new’;
d=“NEW”;/*字符型变量输入时是什么，值就是什么*/
f=“NeW“;
name1=“Tom’s”;
name2=‘Tom’’s’;
date1=‘1jan2006’d;/*直接定义日期*/
date2=‘01jan04’d;
time1=‘9:
25’t;/*直接定义时间*/
time2=‘9:
25:
19’t;
dtime=‘18jan2003:
9:
27:
05am’dt;/*定义日期时间*/
ifbegin=’01may04:
9:
30:
00’dtthenend=”31dec90:
5:
00:
00”dt;
run;
procprint;run;
/*自定义变量*/
/*产生一个取值为1到100的变量*/
dataint;
doi=1to100;/*此处do…end为循环语句,i为循环指标，可取任何符号*/
a=i;
output;
end;
run;
procprint;run;
/*简洁版*/
dataint;
doa=1to100;
output;
end;
run;
procprint;run;
/*通过各种运算定义变量*/
SAS算子
✧用于比较：
=（EQ）等于，^=（NE）不等于，~=（NE）不等于，>（GT）大于，<（LT）小于，>=（GE）大于等于，<=（LE）小于等于
✧算数运算：
+加法，-减法，*乘法，/除法，**幂次
✧逻辑运算:
&（AND）和,|（OR）或，~（NOT）非，^（NOT）非
/*旅游数据*/
datatravel;
inputcountry$nightsaircostlandcostvendor$;
cards;
France8793575Major
Spain10805510Hispania
India10.489Royal
Peru7722590Mundial
;
run;
datanewair;
settravel;
lengthremarks$30;/*remark的值会很长。
所以事先设定长度*/
newacost=aircost+10;/*计算新的机票价格*/
ifvendor='Hispania'thenremarks='Bonusfor10+people';
elseifvendor='Mundial'thenremarks='Bonuspoints';/*对航线加说明*/
elseifvendor='Major'thenremarks='Discountfor30+people';
ifcountry='Peru'thendelete;/*去掉一个已经停飞的航线*/
run;
/*另一个旅游数据*/
dataarttour;
inputcity$1-9nights11landcost13-16events18
describe$20-36guide$38-45backup$47-54;
cards;
Rome375074M,3GD'AmicoTorres
Paris8168065M,1otherLucasLucas
London6123053M,2GWilsonLucas
NewYork6.85M,1G2otherLucasD'Amico
Madrid337053M,2otherTorresD'Amico
Amsterdam458063M,3GVandever
;
run;
datanewarttour;
setarttour;
ifcity='Rome'thenlandcost=landcost+30;/*修改旅游信息*/
ifevents>nightsthencalendar='Checkschedule';
elsecalendar='Noproblems';
ifguide='Lucas'andnights>7thenguide='Torres';
iflandcost>=1500thenprice='High';/*定义新变量price*/
elseiflandcost>=700thenprice='Medium';
elseprice='Low';
ifevents/nights>2thenpace='Toofast';
elsepace='OK';
if（city='Paris'orcity='Rome'）and（guide='Lucas'or
guide="D'Amico"）thentopic='Arthistory';
iflandcostgt1500orlandcost/nightsgt200thenlevel='Deluxe';
ifupcase（city）='MADRID'thenguide='Duncan';/*使用upcase函数*/
ifguide<=:
'L'thengroup='A-L';/*如果guide以L开头*/
elsegroup='M-Z';
/*使用index函数,注意if的用法*/
ifindex（describe,'other'）thenotherev='Yes';
elseotherev='No';
run;
procprintdata=newarttour;title'Tourinformation';run;
/*通过SAS函数定义变量*/
/*函数ranuni生成服从均匀分布的随机数*/
datarandom（drop=i）;
doi=1to3;
Y1=ranuni（1298573062）;
Y2=ranuni（447801538）;
Y3=ranuni（631280）;
output;
end;
procprintdata=random;title“Asinglestreamacrossmultiplevariables”;
run;
/*函数anyupper给出一个字符串中开始位置后的第一个大写字母的位置*/
data_null_;
string=‘Next=_n_+12E3’;
j=0;
dountil（j=0）;
j=anyupper（string,j+1）;
ifj=0thenput+3“That’sall”;
elsedo;
c=substr（string,j,1）;
put+3j=c=;
end;
end;
run;
/*Cakedata*/
datacake;
inputLastName$1-12Age13-14PresentScore16-17
TasteScore19-20Flavor$23-32Layers34;
datalines;
Orlando279380Vanilla1
Ramey328472Rum2
Goldston466875Vanilla1
Roe387973Vanilla2
Larsen237784Chocolate.
Davis518691Spice3
Strickland198279Chocolate1
Nguyen577784Vanilla.
Hildenbrand338183Chocolate1
Byron627287Vanilla2
Sanders265679Chocolate1
Jaeger436674Chocolate2
Davis286975Vanilla1
Conrad698594Chocolate2
Walters556772Spice2
Rossburger287881Chocolate2
Matthew428192Spice2
Becker366283Chocolate1
Anderson278785Chocolate1
Merritt627384Vanilla2
;
PROCPRINT;TITLE"DataCake";RUN;
datacake;
setcake;
score=sum（PresentScore,TasteScore）*sqrt（layers）;
score=round（score,.1）;
Lastname=left（upcase（lastname））;
Lastname1='Mr.orMs.'||Lastname;
DropPresentScoreTasteScoreLayers;
labelscore="FinalScore"
flavor="CakeFlavor";
run;
Procprint;run;
二．循环语句
/*vacationdata*/
datavacation;
inputidyears;
datalines;
12
23
35
46
67
78
;
run;
datavacation;
setvacation;
ifyears>5thendo;
weeks=3+year/5;
group='Qualified';
end;/*这里doend不是循环语句，是大括号*/
elsedo;
yrsleft=5-years;
group='NotQualified';
end;
run;
procprint;run;
/*Investmentdata*/
datainvestment;
begin='01JAN1990'd;
end='31DEC2009'd;
capital=0;
doyear=year（begin）toyear（end）;/*函数year给出时间变量中的年的值*/
capital=Capital+2000+.07*（Capital+2000）;
output;
end;
put'ThenumberofDATAstepiterationsis'_n_;
run;
procprintdata=investment;formatCapitaldollar12.2;run;
datainvestment;
begin='01JAN1990'd;
end='31DEC2009'd;
doyear=year（begin）toyear（end）;
Capital+2000+.07*（Capital+2000）;/*自动给出初值和更新变量值*/
output;
end;
put'ThenumberofDATAstepiterationsis'_n_;
run;
procprintdata=investment;formatCapitaldollar12.2;run;
/*anotherexample*/
dataa;
inputx;
datalines;
5
000
2500
;
run;
datab;
seta;
a=normal（0）;/*normal函数产生一个正态随机数，0表示取现在的时间作为seed*/
exit=10;
doi=1toexit;
y=x*normal（0）;
ify>25theni=exit;
output;
end;
run;
procprint;run;
三．SAS的分组-By语句
/*by语句的定义*/
datazipcodes;
inputstate$1-2city$5-14street$15-25zipcode30-34;
cards;
AZTucsonDomenicLn85730
FLMiamiTradeAve33133
FLMiamiThomasAve33133
FLLakelandFrenchAve33801
FLMiamiNerviaSt33146
AZTucsonGleesonPl85730
FLMiamiRicest33133
FLMiamiCorsicaSt33146
FLLakelandEgretDr33809
FLMiamiSurreyDr33133
;
run;
procprint;run;
/*将同一个州的观测值放一起,by语句的变量为state*/
procsortdata=zipcodesout=state;bystate;run;
procprint;run;
/*进一步将州内的城市排序,by语句的变量为statecity*/
procsortdata=zipcodesout=city;bystatecity;run;
procprint;run;
/*所以变量都排序，by语句的变量为statecityzipcodestreet*/
procsortdata=zipcodesout=city;bystatecityzipcodestreet;run;
procprint;varstatecityzipcodestreet;run;
/*将state按逆序排列*/
procsortdata=zipcodesout=city;
bydescendingstatecityzipcodestreet;
run;
procprint;varstatecityzipcodestreet;run;
/*将state和city都按逆序排列*/
procsortdata=zipcodesout=city;
bydescendingstatedescendingcityzipcodestreet;
run;
procprint;varstatecityzipcodestreet;run;
/*将数据按zipcode分组打印*/
procprint;varstatecityzipcodestreet;bystatecityzipcode;run;
/*procsort的另一个选项*/
datadupobs;
inputcountry$1-11tourtype$13-24nightslandcostvendor$;
cards;
Spainarchitecture10510World
Japanarchitecture8720Express
Switzerlandscenery9734World
Switzerlandscenery9734World
Francearchitecture8575World
Irelandscenery7558Express
NewZealandscenery161489Southsea
Italyarchitecture8468Express
Greecescenery12698Express
;
run;
procprintdata=dupobs;title'DataSetDUPOBS';run;
procsortdata=dupobsout=fixednoduplicates;bycountry;run;
procprintdata=fixed;
title'RemovingaDuplicateObservationwithPROCSORT';
run;
/*确定由by语句定义的组的边界*/
/*当使用procsort,系统会自动生成first.varible和last.variable*/
procsortdata=zipcodes;bystatecity;run;
data_null_;/*不生成数据，但使用datastep*/
setzipcodes;
bystatecity;
if_N_=1thenput"Groupedbystatecity";/*将信息写到log窗口*/
put;
put_N_=state=city=first.state=last.state=first.city=last.city=;
put;/*打一个空行*/
run;
/*计算每组的工资和*/
datasalaries;
inputDepartment$Name$WageCategory$WageRate;
datalines;
BADCarolSalaried20000
BADElizabethSalaried5000
BADLindaSalaried7000
BADThomasSalaried9000
BADLynneHourly230
DDGJasonHourly200
DDGPaulSalaried4000
PPDKevinSalaried5500
PPDAmberHourly150
PPDTinaSalaried13000
STDHelenHourly200
STDJimSalaried8000
;
procprintdata=salaries;run;
procsortdata=salariesout=temp;byDepartment;run;
databudget（keep=DepartmentyearlywagePayroll）;
settemp;
byDepartment;
retainpayroll;
ifWageCategory="Salaried"thenYearlyWage=WageRate*12;
elseifWageCategory="Hourly"thenYearlyWage=WageRate*2000;
/*SASsetsFIRST.variableto1ifthisisanewdepartmentintheBYgroup*/
iffirst.DepartmentthenPayroll=0;
payroll=payroll+YearlyWage;
/*SASsetsLAST.variableto1ifthisisthelastdepartmentintheBYgroup*/
/*iflast.Department;*//*只需取一组中最后一个观测值*/
run;
procprintdata=budget;
formatPayrolldollar10.;
title“AnnualPayrollbyDepartment”;
run;
/*求身高的平方和*/
DATADEMO;
INPUTGENDER$HEIGHTWEIGHTAGE;
DATALINES;
M6815523
F619920
F6311521
M7020545
M69170.
F6512530
M7222048
;
Run;
PROCPRINT;RUN:
dataa;
setDEMOend=done;/*done是一个指示变量，对最后一个观测值为1，其他为0*/
retainNsum0;
N=N+1;
Sum=sum+height**2;
ifdonethendo;
mean=sum/N;
output;
end;
run;
procprint;run;
datadonkey;
inputyear1-4pres$6-29vicepres$31-55result$60-64;
cards;
1920JamesM.CoxFranklinD.Rooseveltlost
1924JohnW.DavisCharlesW.Bryanlost
1928AlfredE.SmithJosephT.Robinsonlost
1932FranklinD.RooseveltJohnN.Garnerwon
1936FranklinD.RooseveltJohnN.Garnerwon
1940FranklinD.RooseveltHenryA.Wallacewon
1944FranklinD.RooseveltHarryS.Trumanwon
1948HarryS.TrumanAlbenW.Barkleywon
1952AdlaiE.StevensonJohnJ.Sparkmanlost
1956AdlaiE.StevensonEstesKefauverl
升级会员 