the r book graphics

《the r book graphics》由会员上传分享，免费在线阅读，更多相关内容在工程资料-天天文库。

1、5GraphicsProducinghigh-qualitygraphicsisoneofthemainreasonsfordoingstatisticalcomputing.Theparticularplotfunctionyouneedwilldependonthenumberofvariablesyouwanttoplotandthepatternyouwishtohighlight.Theplottingfunctionsinthischapteraredealtwithunderfourheadings:
plotswithtw

2、ovariables;
plotsforasinglesample;
multivariateplots;
specialplotsforparticularpurposes.ChangestothedetailedlookofthegraphsaredealtwithinChapter29.5.1PlotswithtwovariablesWithtwovariables(typicallytheresponsevariableontheyaxisandtheexplanatoryvariableonthexaxis),thekindof

3、plotyoushouldproducedependsuponthenatureofyourexplanatoryvariable.Whentheexplanatoryvariableisacontinuousvariable,suchaslengthorweightoraltitude,thentheappropriateplotisascatterplot.Incaseswheretheexplanatoryvariableiscategorical,suchasgenotypeorcolourorgender,thentheappr

4、opriateplotiseitherabox-and-whiskerplot(whenyouwanttoshowthescatterintherawdata)orabarplot(whenyouwanttoemphasizetheeffectsizes).ThemostfrequentlyusedplottingfunctionsfortwovariablesinRarethefollowing:
plot(x,y)scatterplotofyagainstx;
plot(factor,y)box-and-whiskerplotofya

5、teachfactorlevel;
barplot(y)heightsfromavectorofyvalues(onebarperfactorlevel).TheRBook,SecondEdition.MichaelJ.Crawley.©2013JohnWiley&Sons,Ltd.Published2013byJohnWiley&Sons,Ltd.190THERBOOK5.2Plottingwithtwocontinuousexplanatoryvariables:ScatterplotsTheplotfunctiondrawsaxes

6、andaddsascatterplotofpoints.Twoextrafunctions,pointsandlines,addextrapointsorlinestoanexistingplot.Therearetwowaysofspecifyingplot,pointsandlinesandyoushouldchoosewhicheveryouprefer:
Cartesianplot(x,y)
formulaplot(y~x)Theadvantageoftheformula-basedplotisthattheplotfunctio

7、nandthemodelfitlookandfeelthesame(responsevariable,tilde,explanatoryvariable).IfyouuseCartesianplots(eastingsfirst,thennorthings,likethegridreferenceonamap)thentheplothas‘xtheny’whilethemodelhas‘ythenx’.Atitsmostbasic,theplotfunctionneedsonlytwoarguments:firstthenameoftheexp

8、lanatoryvariable(xinthiscase),andsecondthenameoftheresponsevariable(yinthiscase):plot(x,y).Theda