研究生SAS教程第五次上机.ppt

《研究生SAS教程第五次上机.ppt》由会员上传分享，免费在线阅读，更多相关内容在教育资源-天天文库。

1、dataex;doa=1to3;dob=1to3;doi=1to5;inputx@@;output;end;end;end;cards;0.70.60.90.50.60.90.90.71.10.70.80.60.91.00.81.21.41.61.21.51.10.91.31.21.01.51.40.91.31.60.60.60.80.90.70.50.80.91.00.60.61.20.80.91.0;procanova;classab;modelx=ab(a);meansa/lsdcldiff;run;1.p

2、rocanova方差分析——二级套方差分析DependentVariable:xSumofSourceDFSquaresMeanSquareFValuePr>FModel82.800444440.3500555610.07<.0001Error361.252000000.03477778CorrectedTotal444.05244444R-SquareCoeffVarRootMSExMean0.69105119.653300.1864880.948889SourceDFAnovaSSMeanSquareFVal

3、uePr>Fa22.369777781.1848888934.07<.0001b(a)60.430666670.071777782.060.0821tTests(LSD)forxNOTE:ThistestcontrolstheTypeIcomparisonwiseerrorrate,nottheexperimentwiseerrorrate.Alpha0.05ErrorDegreesofFreedom36ErrorMeanSquare0.034778CriticalValueoft2.02809LeastSign

4、ificantDifference0.1381Comparisonssignificantatthe0.05levelareindicatedby***.DifferenceaBetween95%ConfidenceComparisonMeansLimits2-30.480000.341900.61810***2-10.493330.355230.63144***3-2-0.48000-0.61810-0.34190***3-10.01333-0.124770.151441-2-0.49333-0.63144-0

5、.35523***1-3-0.01333-0.151440.12477dataex;inputxy@@;cards;1.54.81.85.72.4738.33.510.93.912.44.413.14.813.6515.32;procgplot;ploty*x;/*以y为纵坐标，以x为横坐标*/symboli=rlv=dot;/*i=rl表示画回归直线*//*v=dot表示观测值对应的点标记为小圆点*/procreg;modely=x/cli;run;/*y=x表示以y为因变量，以x为自变量Cli各个观测值的95

6、%置信上下界Clm每一个因变量Y的期望值的95%置信上下界*/2.procreg回归分析dataex;inputxy@@;cards;0.1420.1143.5.1245.1345.5.1547.5.1649.1753.1850.2055.2155.2360.14.;/*（需预测的y值用缺省符号.表示）*/procreg;modely=x/cli;outputout=a1predicted=pred;/*显示各预测值*/procprintdata=a1;run;TheREGProcedureModel:MODEL

7、1DependentVariable:yAnalysisofVarianceSumofMeanSourceDFSquaresSquareFValuePr>FModel1112.48368112.48368387.52<.0001Error72.031880.29027CorrectedTotal8114.51556RootMSE0.53877R-Square0.9823DependentMean10.12222AdjR-Sq0.9797CoeffVar5.32260ParameterEstimatesParame

8、terStandardVariableDFEstimateErrortValuePr>

9、t

10、Intercept10.256950.532350.480.6441x12.930280.1488619.69<.0001方差的估计的平方根称为标准误Model:MODEL1DependentVariable:yOutputStatisticsDepVarPredictedStdE