Matlab统计工具箱

《Matlab统计工具箱》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、Matlab统计工具箱GettingStartedToolboxfunctionsandGUIsOrganizingDataDataarraysandgroupsDescriptiveStatisticsDatasummariesStatisticalVisualizationDatapatternsandtrendsProbabilityDistributionsModelingdatafrequencyRandomNumberGenerationTopicsinrandomnumbergenerationHypothesisTestsInferencesfromdata

2、AnalysisofVarianceModelingdatavarianceRegressionAnalysisContinuousdatamodelsMultivariateMethodsVisualizationandreductionClusterAnalysisIdentifyingdatacategoriesClassificationCategoricaldatamodelsMarkovModelsStochasticdatamodelsDesignofExperimentsSystematicdatacollectionStatisticalProcessCo

3、ntrolProductionmonitoring方差分析ANOVAOperationsanova1One-wayanalysisofvarianceanova2Two-wayanalysisofvarianceanovanN-wayanalysisofvarianceaoctoolInteractiveanalysisofcovariancedummyvarCreatedummyvariablesfriedmanFriedman'stestkruskalwallisKruskal-Wallistest22manova1One-waymultivariateanalysis

4、ofvariancemanovaclusterDendrogramofgroupmeanclustersfollowingMANOVAmultcompareMultiplecomparisontest方差分析模型方差分析是分析试验（或观测）数据的一种统计方法。在工农业生产和科学研究中，经常要分析各种因素及因素之间的交互作用对研究对象某些指标值的影响。这时需要用到方差分析。利用方差分析，我们能推断哪些因素对所考察指标的影响是显著的，哪些是不显著的。方差分析包括了：单因素方差分析、双因素方差分析和多因素方差分析。单因素方差分析单因素方差分析问题：某一因素（A）对结果（）的影响分

5、析。一般情况下，假设因素A有个不同的水平。单因素方差分析模型如下：其中表示在第组（水平）第个观察值，为总的（的）均值，为第组所的均值且满足，表示第组第个样本的误差，为第组的数据个数且。方差分析的目的是要基于数据分析确定这组数据的均值有没有显著性差异，即假设检验问题：v.s.不全相等。（1）原假设表明因素A对的值没有影响。22格式p=anova1(X)p=anova1(X,group)p=anova1(X,group,displayopt)[p,table]=anova1(...)[p,table,stats]=anova1(...)实例1CreateXwithcolumns

6、thatareconstantsplusrandomnormaldisturbanceswithmeanzeroandstandarddeviationone:X=meshgrid(1:5)X=1234512345123451234512345X=X+normrnd(0,1,5,5)X=1.35502.06622.46885.94475.48972.06931.76111.48644.88266.32222.19190.72763.19054.87684.68412.76201.81793.95064.46784.9291-0.36261.16853.57422.19455

7、.9465Performone-wayANOVA:p=anova1(X)p=227.9370e-006theverysmallpvalueindicatesthatdifferencesbetweencolumnmeansarehighlysignificant.Theprobabilityofthisoutcomeunderthenullhypothesis(thatsamplesdrawnfromthesamepopulationwouldhavemeansdifferingbytheamountsseenin