SASbase70答案詳解綜述

SasCrackmanpracticeContents1）first.variablelast.variable 1Thefollowingexampleisbetter: 2Equivalentcoding: 32Readrawdata@missoverdsmdlm 7Infiledsd 7Infiledlm 9Infilemissover 10Input@ 103Libnamedatasetexcel‘’ 114FunctionMDYcharacterversusnumericvariable 135Infileobs 136Keepstatementandkeepoption 137ODSexcelfile 138first.variablelastvariable 139Thevalueisnotintherangeprocformatvalue 1310Charactervariablebeingautomaticallyconvertedtothenumericvariable 1311Theoutputoftheprocmeans 1312libnamedataset‘Dallas.xls’; 1313Twowaystodefinearrays 13Datadatasetname/debug; 1315._error_willbeautomaticallydropped 1316Day()function 1317Ifthenstatement.Twostatementsafterthen 1319WHERE=datasetoptiondonotperformanyautomaticconversionincomparisons 1320subsettingif 1321Proccontentswork.all; 1322libnameprog2'C:\Users\dude\Desktop\prog2'; 1324infileDSD 1325firstobs 1326Procsort 1327Procsortbydescending 1328Mergeandrename 1329Trailing@ 1330Executionerror 1331Missingdataattheendoftherow 1332Sumstatement 1333characterformatiscase-sensitive 1334procformatwillgenerateafilestoredinthecatalog. 1335(tough)Writetomultipledatasets 1336Informatmmddyy10. 1338First.variable 1339Arraydefinesnewcharactervariables 1340dowhilevs.dountil 1341CATXfunction 1342Findfunction.‘I’and‘T’asmodifiersand5asthestartingposition. 1343Putfunctiongivesyoucharacter;inputgivesyounumbers 13TRANWRDfunction 13inputfunction 13Putfunction 13CHARFunction 13Syntax 13Arguments 1344.don’tforgetthesemicolons 1345.odscsvallfile='c:\test.cvs'; 1346mean(ofRev:)Usecolonstorepresentalistofvariables. 1347Tablesvariable1*variable2/nocolnorow 1348Columninput 1349Anarraycannotbereferencedonakeep=datasetoption 1350Interleavingdatasetswithbystatement:whichvariablecomesfirst. 1351Proccontentsdata=_all_; 1352Merge.Indatasetoption 1353Retainstatementinitializesthevaluetomissingbydefault. 1354retainCity'BeverlyHills'; 1355.Sasdateconstant:noblankbefored. 1356.Tablesvarible1*variable2 1357Sasdatefunction 1358Procformat;valuename(noperiod) 1359Delimitedrawdatafileandtruncover 1360Dropoptioninthedatastep 1361ifAgenotin(15,16)then 1362odshtmlfile='sales.html' 1363Dountil 1365Permanentlyassignlabelsinthedatastep 1366doloop 1367Seeamissingvalueasnegativeinfinity 1368Charactervariablevs.numericvariable. 1369whereCodelike'E_U%'; 1370Datastepdropoption 131）first.variablelast.variableOn2010/11/20,in跟crackman做sasbase認證試題,bycrackmanBasically,thismeans1.ThefollowingSASprogramissubmitted:dataWORK.TOTAL;setWORK.SALARY;byDepartmentGender;ifFirst.<_insert_code_>thenPayroll=0;Payroll+Wagerate;ifLast.<_insert_code_>;run;TheSASdatasetWORK.SALARYiscurrentlyorderedbyGenderwithinDepartment.WhichinsertedcodewillaccumulatesubtotalsforeachGenderwithinDepartment?A.GenderB.DepartmentC.GenderDepartmentD.DepartmentGenderAnswer:AComment:Weknowthatfirst.variableislikeadummyvariablethatidentifiesthefirstvariableas1andothersas0.Thefollowingexampleisbetter:datatemp;inputgroupx;cards;1231341.145278292245289234276331423412;run;/**************************************************Theautomaticvariablesfirst.groupandlast.grouparenotsavedwiththedataset.Herewewritethemtodatasetvariablestoshowtheircontents.**************************************************/datanew;settemp;bygroup;first=first.group;last=last.group;run;procprint;title'Rawdataalongwithfirst.groupandlast.group';run;/**************************************************AcommontaskindatacleaningistoidentifyobservationswithaduplicateIDnumber.IfwesetthedatasetbyID,thentheobservationswhicharenotduplicatedwillbeboththefirstandthelastwiththatIDnumber.Wecanthereforewriteanyobservationswhicharenotbothfirst.idandlast.idtoaseparatedatasetandexaminethem.**************************************************/datasingledup;settemp;bygroup;iffirst.groupandlast.groupthenoutputsingle;elseoutputdup;run;/**************************************************Wemayalsowanttododatasetprocessingwithineachbygroup.InthisexampleweconstructthecumulativesumofthevariableXwithineachgroup.**************************************************/datacusum(keep=groupsum);settemp;bygroup;iffirst.groupthensum=0;sum+x;iflast.groupthenoutput;run;procprintdata=cusumnoobs;title'SumofXwithineachgroup';run;SumofXwithineachgroup08:30Wednesday,May23,20127groupsum11022414331435Equivalentcoding:/**************************************************Asanaside,ifyousimplywantthesumofXwithineachgroup,oneofthemanywayofobtainingthisiswithPROCPRINT.**************************************************/procprintdata=temp;title'AlldatawithXsummedwithineachgroup';bygroup;sumx;sumbygroup;run;procmeansdata=tempsum;bygroup;run;AlldatawithXsummedwithineachgroup808:30Wednesday,May23,2012group=1Obsx1232343.445group102group=2Obsx5786927458899341076group414group=3Obsx1131group=4Obsx12231312group35===582

AlldatawithXsummedwithineachgroup908:30Wednesday,May23,2012group=1TheMEANSProcedureAnalysisVariable:xSum????????????102.0000000????????????group=2AnalysisVariable:xSum????????????414.0000000????????????group=3AnalysisVariable:xSum????????????31.0000000????????????group=4AnalysisVariable:xSum????????????35.0000000????????????本題考察的是：first.var和last.var這兩個知識點。本題的意思，數(shù)據(jù)集SALAR已經(jīng)按照department和gender排序，現(xiàn)在計算每一個department下每一個gender的某一個變量值的累積和。根據(jù)本題的意思，寫了一個模擬程序：datacrackman;inputdepartment$gender$salary@;datalines;marketf6000marketm5000marketf5500marketm8000marketf6000marketm7000salesf6000selesm4000salesf6000selesm4000salesf6000selesm4000;procsortdata=crackman;bydepartmentgender;run;dataresult;setcrackman;/*1*/bydepartmentgender;iffirst.genderthensubtotal=salary;/*2*/elsesubtotal+salary;/*3*/iflast.gender;/*4*/run;/*5*/這個DATARESULT部分程序執(zhí)行的過程是這樣（不是編譯過程）：/*我把執(zhí)行的語句分為標記12345*/1.1–2–4–1（先讀數(shù)據(jù)—判斷first.gender是否為1，第一個肯定是1——所以直接跳過else到iflast.gender，last.gender=0所以—setcrackman，繼續(xù)讀數(shù)據(jù))2.1–2–3–4–13.1–2–3–4–5–1(因為是第三個了，所以last.gender=1然后到run,輸出到result.后面都是一樣。這里第三條語句其實===subtotal=subtotal+salary，只是這里簡寫了，體現(xiàn)程序的簡潔性，但是對于入門同學來說增加了可讀性的難度。2Readrawdata@missoverdsmdlm2.GiventhefollowingrawdatarecordsinTEXTFILE.TXT:

—-|—-10—|—-20—|—-30

John,FEB,13,25,14,27,Final

John,MAR,26,17,29,11,23,Current

Tina,FEB,15,18,12,13,Final

Tina,MAR,29,14,19,27,20,Current

Thefollowingoutputisdesired:

Obs

Name

Month

Status

Week1

Week2

Week3

Week4

Week5

1

John

FEB

Final

$13

$25

$14

$27

.

2

John

MAR

Current

$26

$17

$29

$11

$23

3

Tina

FEB

Final

$15

$18

$12

$13

.

4

Tina

MAR

Current

$29

$14

$19

$27

$20

WhichSASprogramcorrectlyproducesthedesiredoutput?Infiledsd

A.

dataWORK.NUMBERS;

lengthName$4Month$3Status$7;

\o"谷歌搜索tag結(jié)果列表"infile‘TEXTFILE.TXT’dsd;

\o"谷歌搜索tag結(jié)果列表"inputName$Month$;

ifMonth=’FEB’then\o"谷歌搜索tag結(jié)果列表"inputWeek1Week2Week3Week4Status$;

elseifMonth=’MAR’then\o"谷歌搜索tag結(jié)果列表"inputWeek1Week2Week3Week4Week5Status$;

formatWeek1-Week5dollar6.;

run;

procprintdata=WORK.NUMBERS;

run;

A.DSD：規(guī)定若一個數(shù)據(jù)是由引號，那么SAS認為引號內(nèi)的逗號也是屬于字符數(shù)據(jù)；設(shè)定默認分隔符為逗號；連續(xù)兩個分隔符之間的數(shù)據(jù)位缺失值；讀入數(shù)據(jù)時去掉引號。

input語句在讀入觀測時，先從第一行“John,FEB,13,25,14,27,Final”讀取數(shù)據(jù)“John,FEB”就轉(zhuǎn)入到了第二行“John,MAR,26,17,29,11,23,Current”，那么因為month=”FEB”是true

所以接著在第二行“John,MAR,26,17,29,11,23,Current”，讀入week1week2，但是johnmar都是字符型，而week1week2默認為數(shù)值型，所以就是缺失值。繼續(xù)讀入week3week4也就是2617

接著就是status是29。讀完之后回到infile，回到input,從第三行開始讀”Tina,FEB,15,18,12,13,Final”,讀入namemonth之后，轉(zhuǎn)達到第四行“Tina,MAR,29,14,19,27,20,Current”。

infiledlm

B.

dataWORK.NUMBERS;

lengthName$4Month$3Status$7;

infile‘TEXTFILE.TXT’dlm=’,’missover;

\o"谷歌搜索tag結(jié)果列表"inputName$Month$;

ifMonth=’FEB’then\o"谷歌搜索tag結(jié)果列表"inputWeek1Week2Week3Week4Status$;

elseifMonth=’MAR’then\o"谷歌搜索tag結(jié)果列表"inputWeek1Week2Week3Week4Week5Status$;

formatWeek1-Week5dollar6.;

run;

procprintdata=WORK.NUMBERS;

run;B.MISSOVER：組織INPUT從下一個數(shù)據(jù)行中讀入數(shù)據(jù)，未賦值的變量就是缺失值。

其實從A看出，因為第一次INPUT只讀入第一數(shù)據(jù)行兩個值，指針轉(zhuǎn)入到第二行了。那么MISSOVER是否可以阻止呢？肯定不能阻止，因為INPUT第一次就只讀入兩個數(shù)據(jù)，而且第一行的數(shù)據(jù)也是大于2個的，所以根本就無需阻止。自己讀完兩個數(shù)據(jù)之后自動轉(zhuǎn)向第二行。不能讓第一次和第二次的input指針保留在同一行中。InfiledlmInfilemissoverExample2:HandlingMissingValuesandShortRecordswithListInputThisexampledemonstrateshowtopreventmissingvaluesfromcausingproblemswhenyoureadthedatawithlistinput.Somedatalinesinthisexamplecontainfewerthanfivetemperaturevalues.UsetheMISSOVERoptionsothatthesevaluesaresettomissing.dataweather;infiledatalinesmissover;inputtemp1-temp5;datalines;97.998.198.398.699.299.198.597.596.297.398.397.696.5;SASreadsthethreevaluesonthefirstdatalineasthevaluesofTEMP1,TEMP2,andTEMP3.TheMISSOVERoptioncausesSAStosetthevaluesofTEMP4andTEMP5tomissingforthefirstobservationbecausenovaluesforthosevariablesareinthecurrentinputdatarecord.WhenyouomittheMISSOVERoptionoruseFLOWOVER,SASmovestheinputpointertoline2andreadsvaluesforTEMP4andTEMP5.ThenexttimetheDATAstepexecutes,SASreadsanewlinewhich,inthiscase,isline3.ThismessageappearsintheSASlog:NOTE:SASwenttoanewlinewhenINPUTstatementreachedpasttheendofaline.

C.

dataWORK.NUMBERS;

lengthName$4Month$3Status$7;

infile‘TEXTFILE.TXT’dlm=’,';

\o"谷歌搜索tag結(jié)果列表"inputName$Month$@;

ifMonth=’FEB’theninputWeek1Week2Week3Week4Status$;

elseifMonth=’MAR’theninputWeek1Week2Week3Week4Week5Status$;

formatWeek1-Week5dollar6.;

run;

procprintdata=WORK.NUMBERS;

run;Input@

Sointhefollowingexample,@willholdthelineandaskSASnottomovetothenextline.

D.

dataWORK.NUMBERS;

lengthName$4Month$3Status$7;

infile‘TEXTFILE.TXT’dsd@;

inputName$Month$;

ifMonth=’FEB’theninputWeek1Week2Week3Week4Status$;

elseifMonth=’MAR’theninputWeek1Week2Week3Week4Week5Status$;

formatWeek1-Week5dollar6.;

run;

procprintdata=WORK.NUMBERS;

run;3Libnamedatasetexcel‘’3.TheExcelworkbookREGIONS.XLScontainsthefollowingfourworksheets:

EAST

WEST

NORTH

SOUTH

Thefollowingprogramissubmitted:

libnameMYXLS‘regions.xls’;

WhichPROCPRINTstepcorrectlydisplaystheNORTHworksheet?

A.procprintdata=MYXLS.NORTH;run;

B.procprintdata=MYXLS.NORTH$;run;

C.procprintdata=MYXLS.’NORTH’e;run;

D.procprintdata=MYXLS.’NORTH$’n;run;Disthecorrectone.Basicallytheabovecodeestablishestheexcellibnameandthenprintoneoftheworksheet.Thekeyistounderstandthatnorth$istheactualexcelname,thoughwecannotsee$.However,itisnotalegitimatesasname.That’swhyweputtheminquotationmarks,followedbyadollarsign.Dollarsignshouldalsobetheinthequotationmarks.Thequotationmarksshouldbefollowedimmediatelybyn,withoutblanks.4FunctionMDYcharacterversusnumericvariable4.ThefollowingSASprogramissubmitted:

dataWORK.DATE_INFO;

Day=”01″;

Yr=1960;

X=mdy(Day,01,Yr);

run;

WhatisthevalueofthevariableX?

A.thenumericvalue0

B.thecharactervalue“01011960″

C.amissingvalueduetosyntaxerrors

D.thestepwillnotcompilebecauseofthecharacterargumentinthemdyfunction.Firstofall,noticethatthedayhereactuallymeansthemonth.Moreimportantly,weshouldactuallydothecodinglikethis:dataWORK.DATE_INFO;Day=01;Yr=1960;X=mdy(Day,01,Yr);run;Intheabovecode,dayvariableisproperlytreatedasanumericvariable.dataWORK.DATE_INFO;Day='01';Yr=1960;X=mdy(Day,01,Yr);run;However,theabovecodestillworks,becauseSASconvertedit.Checkoutthelog:NOTE:Charactervalueshavebeenconvertedtonumericvaluesattheplacesgivenby:(Line):(Column).178:12Answer:A

這里考察的是日期在SAS中如何儲存的。

SAS日期格式都是以數(shù)值型存在的。而且SAS根據(jù)1960年1月1號作為基礎(chǔ)轉(zhuǎn)化目標日期為數(shù)值型數(shù)據(jù)存儲。

X=MDY(DAY,01，YR)編譯之后就是X=MDY(’01′,01,1960)，系統(tǒng)自動將字符型的數(shù)字‘01’轉(zhuǎn)化為數(shù)字1，所以結(jié)果1960年1月1號，所以為0。5Infileobs5.Whichstatementspecifiesthatrecords1through10aretobereadfromtherawdatafilecustomer.txt?

A.\o"谷歌搜索tag結(jié)果列表"infile‘customer.txt’1-10;

B.\o"谷歌搜索tag結(jié)果列表"input‘customer.txt’stop@10;

C.\o"查看infile中的全部文章"infile‘customer.txt’\o"查看obs=中的全部文章"obs=10;

D.\o"谷歌搜索tag結(jié)果列表"input‘customer.txt’stop=10;

Answer:C

這道題考察的是INFILE語句后面的參數(shù)意義

前幾道題考過DSDMISSOVERDLM等

此處只有C是對的。OBS=

IndicateswhichlineinyourrawdatafileshouldbetreatedasthelastrecordtobereadbySAS.Thisisagoodoptiontousefortestingyourprogram.Forexample,youmightuseobs=100tojustreadinthefirst100linesofdatawhileyouaretestingyourprogram.Whenyouwanttoreadtheentirefile,youcanremovetheobs=optionentirely.Atypicalinfilestatementforreadingacommadelimitedfilethatcontainsthevariablenamesinthefirstlineofdatawouldbe:Notethatthereisonstopoptionfortheinfileorinputstatement.6Keepstatementandkeepoption6.AfteraSASprogramissubmitted,thefollowingiswrittentotheSASlog:

101

dataWORK.JANUARY;

102

setWORK.ALLYEAR(\o"查看keep中的全部文章"keep=productmonthnum_SoldCost);

103

ifMonth=’Jan’thenoutputWORK.JANUARY;

104

Sales=Cost*Num_Sold;

105

\o"查看keep中的全部文章"keep=ProductSales;

—–

22

ERROR22-322:Syntaxerror,expectingoneofthefollowing:!,!!,&,*,**,+,-,,<=,<>,=,>,>=,

AND,EQ,GE,GT,IN,LE,LT,MAX,MIN,NE,NG,NL,NOTIN,OR,^=,|,||,~=.

106

run;

WhatchangesshouldbemadetotheKEEPstatementtocorrecttheerrorsintheLOG?

A.\o"查看keep中的全部文章"keep=(ProductSales);

B.\o"查看keep中的全部文章"keepProduct,Sales;

C.\o"查看keep中的全部文章"keep=Product,Sales;

D.\o"查看keep中的全部文章"keepProductSales;Answer:D

本道題考察的是KEEP語句的語法問題

在括號內(nèi)就是KEEP=

如果不是就是KEEPThefollowingDATAstepcreatesthesametwodatasetsastheDATAstepinthepreviousexample,butitdoesnotreadthevariableTotalintotheprogramdatavector.ComparetheSETstatementheretotheoneinCreatingMoreThanOneDataSetinaSingleDATAStep.dataservices(keep=ServicesTotalServicesPoliceServicesFireServicesWater_Sewer)admin(keep=AdminTotalAdminLaborAdminSuppliesAdminUtilities);setcity(drop=Total);run;procprintdata=services;title'CityExpenditures:Services';run;procprintdata=admin;title'CityExpenditures:Administration';run;Incontrastwithpreviousexamples,thedatasetoptionsinthisexampleappearinboththeDATAandSETstatements.IntheSETstatement,theDROP=optiondetermineswhichvariablesareomittedfromtheprogramdatavector.IntheDATAstatement,theKEEP=optioncontrolswhichvariablesarewrittenfromtheprogramdatavectortoeachdatasetbeingcreated.Note:

UsingaDROPorKEEPstatementiscomparabletousingaDROP=orKEEP=optionintheDATAstatement.Allvariablesareincludedintheprogramdatavector;theyareexcludedwhentheobservationiswrittenfromtheprogramdatavectortothenewdataset.WhenyoucreatemorethanonedatasetinasingleDATAstep,usingthedatasetoptionsenablesyoutodroporkeepdifferentvariablesineachofthenewdatasets.ADROPorKEEPstatement,ontheotherhand,affectsallofthedatasetsthatarecreated.

7ODSexcelfile7.WhichofthefollowingchoicesisanunacceptableODSdestinationforproducingoutputthatcanbeviewedinMicrosoftExcel?

A.MSOFFICE2K

B.\o"查看EXCEL中的全部文章"EXCELXP

C.\o"查看CSV中的全部文章"CSVALL

D.WINXPAnswer:D

本題考察的是ODSdestination。

答案是D，前三個都可以再\o"查看EXCEL中的全部文章"EXCEL下打開。不過三者之間還是有區(qū)別。

OneofthebenefitsofusingthistagsetisthatitparentsfromtheMSOffice2Ktagset,andthereforetheoutputcancontainbothgraphicsandtables,unliketheExcelXPdestination,whichcannothandlegraphicscurrently.Tagsets.MSOffice2K_xalsoaddsoptionstoperformmanyofthecommontasksinExcel,muchlikeExcelXP.SomethingsthattheExcelXPtagsetdoesverywell,suchascreatingmultipleworksheetsperworkbook,canbedonehereaswell,butnotasdynamically.Distooridiculous.AbetterfalseoptionwouldbeDexcel.HerearethegeneralODSstatementsneededtogenerateXMLoutputthatiscompatiblewithExcel2002andlater:ods_all_close;odstagsets.ExcelXPfile='file-name.xml'style=style-name...;*YourSASprocedurecodehere;odstagsets.ExcelXPclose;8first.variablelastvariable8.TheSASdatasetnamedWORK.SALARYcontains10observationsforeachdepartment,andiscurrentlyorderedbyDepartment.ThefollowingSASprogramissubmitted:

dataWORK.TOTAL;

setWORK.SALARY(keep=DepartmentMonthlyWageRate);

byDepartment;

ifFirst.Department=1thenPayroll=0;

Payroll+(MonthlyWageRate*12);(Thisisthesumstatement)

ifLast.Department=1;(Thisisthesubsettingif.SAScannotgooverthisstatementunlesslast.department=1)

run;

Bysheercoincidence,wejustlearnedthis.Bydepartment;willcreatetwovariablesfirst.departmentandlast.department.payroll=0istheinitialvalueWhichstatementistrue?

A.ThebystatementintheDATAstepcausesasyntaxerror.

B.ThestatementPayroll+(MonthlyWageRate*12);inthedatastepcausesasyntaxerror.

C.ThevaluesofthevariablePayrollrepresentthemonthlytotalforeachdepartmentintheWORK.SALARYdataset.

D.ThevaluesofthevariablePayrollrepresentamonthlytotalforallvaluesofWAGERATEintheWORK.SALARYdataset.9Thevalueisnotintherangeprocformatvalue9.datacourse;

\o"谷歌搜索tag結(jié)果列表"inputexam;

datalines;

50.1

;

run;\o"查看procformat中的全部文章"procformat;

valuescore1–50=‘Fail’

51–100=‘Pass’;

run;procreportdata=coursenowd;

columnexam;

defineexam/displayformat=score.;

run;Whatisthevalueforexam?

A.

Fail

B.

Pass

C.

50.1

D.

NooutputAnswer:C

其實這道題考察的是FORMAT的FUZZ

結(jié)果是50.1，因為50.1不在FORMATVALUE的區(qū)間內(nèi)。如果改成：

\o"查看procformat中的全部文章"procformat;

valuescore（fuzz=.2）1–50=‘Fail’

51–100=‘Pass’;

run;

大家看看！

還有1-50是[1,50]

51-100是[5