常用試驗設(shè)計及結(jié)果分析

實驗八常用試驗設(shè)計及結(jié)果分析實驗內(nèi)容：試驗設(shè)計完全隨機隨機區(qū)組拉丁方套設(shè)計裂區(qū)設(shè)計結(jié)果分析完全隨機隨機區(qū)組拉丁方裂區(qū)試驗一、試驗設(shè)計用四道指令，其格式：PROCPLAN選項串；FACTORS主效果名稱抽樣方式/NOPRINT;OUTPUTOUT=輸出文件名稱

DATA=SAS文件名主效果的抽樣結(jié)果；TREATMENTS其他效果的抽樣方式SEED一個小于（231-1）的整數(shù)或ORDERED主效果組別1、2、…=RANDOM(內(nèi)設(shè)值)或ORDERED或CYCLIC1.完全隨機試驗設(shè)計PROCPLANSEED=27371;factorsunit=12;

treamentstreat=12cyclic(111111222222);OUTPUTOUT=OUTDAT;PROCSROT;BYUNIT;PROCPRINT;RUN;2.隨機區(qū)組試驗設(shè)計PROCPLANSEED=17371;treatmentstmts=5cyclic(12345)1;factorsblock=4variety=5;OUTPUTOUT=OUTDAT;QUIT;PROCTABULATE;CLASSBLOCKVARIETY;VARTMTS;TABLEBLOCK,VARIETY*(TMTS*F=8.)/RTS=8;RUN;TITLE1"CompletelyBlockDesign";

TITLE2"For4Treatments";

3.拉丁方試驗設(shè)計PROCplanseed=37430;factorsrows=4orderedcols=4ordered/NOPRINT;

treatments

tmts=4cyclic;OUTPUTOUT=OUTDATROWS

cvals=("Day1""Day2""Day3""Day4")randomcols

cvals=("Lab1""Lab2""Lab3""Lab4")randomtmts

nvals=(0100200300)random;

QUIT;PROCTABULATE;CLASSROWSCOLS;VARTMTS;TABLEROWS,COLS*(TMTS*F=6.)/RTS=8;RUN;TITLE1"Latinsquaredesign";TITLE2"For4treatments";TITLE1"拉丁方設(shè)計";TITLE2"4*4";PROCplanseed=37430;factorsrows=4orderedcols=4ordered/noprint;treatmentstrt=4cyclic;OUTPUTOUT=GROWSRANDOMCOLSRANDOMTRTRANDOM;PROCPRINT;RUN;4.巢式試驗設(shè)計TITLE1"HierarchicalDesign";TITLE2“FOR3HOUSES4POTSAND3PLANTS";PROCPLANseed=13371;FACTORShouses=3pots=4plant=3;run;5.裂區(qū)試驗設(shè)計TITLE1“SplitPlotDesign";procplanseed=37277;factorsblock=3orderedA=4B=2;run;二、結(jié)果分析

PROCANOVA過程語句CLASS變量名稱串

MODEL因變量名稱串=實驗效果串TESTH=效果名稱E=效果名稱

MEANSDUNCAN,SNK,T/LSDALPHA=0.01,默認值0.05ABCY=ABCY=ABCA*BA*CB*CA*B*CY=AB(A)CLASS必須在MODEL之前，TEST、MEANS可以重復(fù)使用

主要用五道命令1.完全隨機設(shè)計-單因素DATAwheat;Dotreat=1to5;Dor=1to5;INPUThight@@;OUTPUT;END;END;cards;64.665.364.866.065.864.565.364.663.763.967.866.367.166.868.571.872.170.069.171.069.268.269.868.367.5;PROCANOVA;CLASStreat;MODELhight=treat;MEANStreat/duncan;run;2.完全隨機設(shè)計-三因素OPTIONSLINESIZE=76;DATAexample;DOa=1to3;dob=1to4;doc=1to5;dor=1to2;INPUTy@@;OUTPUT;END;END;END;END;cards;;PROCANOVA;CLASSabc;MODELy=abca*ba*cb*ca*b*c;TESTH=aE=a*b;TESTH=cE=b*c;TESTH=a*cE=a*b*c;MEANSa/duncane=a*balpha=0.01;MEANSc/LSDe=b*calpha=0.01;run;3.隨機區(qū)組設(shè)計-單因素OPTIONSLINESIZE=76;DATAwheat;INPUTblockvarietyyield@@;CARDS;111812361331142115302123223023342418253031223430333434183542;;PROCANOVA;CLASSblockvariety;MODELyield=varietyblock;MEANSvariety/duncan;run;4.隨機區(qū)組設(shè)計-兩因素OPTIONSLINESIZE=76;DATAwork;DOblock=1TO2;DOa=1TO4;DOb=1TO4;INPUTenergy@@;OUTPUT;END;END;END;CARDS;;;PROCANOVA;CLASSblockab;MODELenergy=blockaba*b;TESTH=aE=a*b;MEANSa/duncanE=a*b;run;5.正交拉丁方試驗OPTIONSLINESIZE=76;DATAsoybean;INPUTrcvmyield@@;CARDS;;;PROCANOVA;CLASSrcvm;MODELyield=rcvm;MEANSv/duncanalpha=0.01;run;6.套設(shè)計OPTIONSLINESIZE=76;DATAnested;doplant=1to3;doleaf=1to2;dol=1to2;INPUTwt@@;output;end;end;end;CARDS;12.112.112.812.814.414.414.714.523.123.428.128.8;PROCANOVA;CLASSplantleaf;MODELwt=plantleaf(plant);TESTH=plante=leaf(plant);meansplant/lsd;run;

DATAsplit;DOb=1TO4;DOc=1TO4;DOblock=1TO2;INPUTenergy@@;OUTPUT;END;END;END;CARDS;;

PROCANOVA;CLASSblockbc;MODELenergy=blockbcblock*bblock*cb*cblock*b*c;testh=blocke=bl