r - 使用 ggplot 用不同颜色绘制多个数据帧

Question

我对 r 和 ggplot2 都很陌生。我正在尝试使用下面的代码创建一个绘图，其想法是主数据集(dat)包含我想要绘制的数据，并且 dat 中的每个“tp”都被分配了唯一的颜色，以便我可以看到厚与特定的“tp”相关。此代码运行良好，输出符合预期。

#Example
require(ggplot2)

#DataFrame
mytp<-c(80.01,80.01,80.01,80.01,80.01,80.01,80.02,80.02,80.02,80.02,80.02,80.02,80.03,80.03,80.03,80.03,80.03,80.03,80.04,80.04,80.04,80.04,80.04,80.04,83.01,83.01,83.01,83.01,83.01,83.01,83.02,83.02,83.02,83.02,83.02,83.02,83.03,83.03,83.03,83.03,83.03,83.03,83.04,83.04,83.04,83.04,83.04,83.04)
mycompid<-c(4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4)
mytml<-c(80,80,80,80,80,80,80,80,80,80,80,80,80,80,80,80,80,80,80,80,80,80,80,80,83,83,83,83,83,83,83,83,83,83,83,83,83,83,83,83,83,83,83,83,83,83,83,83)
myyears<-c(7.33470226,23.50171116,20.25188227,15.8329911,26.58726899,11.00068446,7.33470226,26.58726899,23.50171116,11.00068446,20.25188227,15.8329911,7.33470226,11.00068446,26.58726899,23.50171116,15.8329911,20.25188227,7.33470226,11.00068446,15.8329911,20.25188227,26.58726899,23.50171116,7.33470226,11.00068446,15.8329911,20.25188227,23.50171116,26.58726899,7.33470226,11.00068446,15.8329911,20.25188227,23.50171116,26.58726899,7.33470226,11.00068446,15.8329911,20.25188227,23.50171116,26.58726899,7.33470226,11.00068446,15.8329911,20.25188227,23.50171116,26.58726899)
mythick<-c(0.29,0.28,0.28,0.28,0.28,0.28,0.29,0.28,0.28,0.28,0.28,0.28,0.28,0.28,0.28,0.28,0.28,0.28,0.27,0.27,0.27,0.27,0.27,0.27,0.28,0.27,0.27,0.27,0.27,0.27,0.28,0.28,0.28,0.28,0.28,0.28,0.3,0.29,0.29,0.28,0.28,0.28,0.3,0.28,0.28,0.27,0.27,0.27)
mycomp<-c("6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE")
myDate<-c("1994-10-01","2010-12-01","2007-08-31","2003-03-31","2013-12-31","1998-05-31","1994-10-01","2013-12-31","2010-12-01","1998-05-31","2007-08-31","2003-03-31","1994-10-01","1998-05-31","2013-12-31","2010-12-01","2003-03-31","2007-08-31","1994-10-01","1998-05-31","2003-03-31","2007-08-31","2013-12-31","2010-12-01","1994-10-01","1998-05-31","2003-03-31","2007-08-31","2010-12-01","2013-12-31","1994-10-01","1998-05-31","2003-03-31","2007-08-31","2010-12-01","2013-12-31","1994-10-01","1998-05-31","2003-03-31","2007-08-31","2010-12-01","2013-12-31","1994-10-01","1998-05-31","2003-03-31","2007-08-31","2010-12-01","2013-12-31")
mytpcolor<-c("gray53","gray53","gray53","gray53","gray53","gray53","lightyellow4","lightyellow4","lightyellow4","lightyellow4","lightyellow4","lightyellow4","salmon2","salmon2","salmon2","salmon2","salmon2","salmon2","pink","pink","pink","pink","pink","pink","gray52","gray52","gray52","gray52","gray52","gray52","mistyrose1","mistyrose1","mistyrose1","mistyrose1","mistyrose1","mistyrose1","dodgerblue","dodgerblue","dodgerblue","dodgerblue","dodgerblue","dodgerblue","violetred1","violetred1","violetred1","violetred1","violetred1","violetred1")

#Primary Data for Plot
dat<-data.frame(tp=mytp, compid=mycompid,       tml=mytml,years=myyears,thick=mythick,comp=mycomp,Date=myDate,tpcolor=mytpcolor)
dat$Date=as.Date(dat$Date)

ldatcompid<-c(4,4,4,4,4,4,4,4,4)
ldatnps<-c(6,6,6,6,6,6,6,6,6)
ldatcomp<-c("PIPE","PIPE","PIPE","PIPE","PIPE","PIPE","PIPE","PIPE","PIPE")
ldatcase<-c("data","data","data","data","data","data","data","data","data")
ldatyear<-c(0,2.5,7.33,11,15.83,19.83,20.25,23.5,26.59)
ldatMean<-c(0.2849,0.2843,0.2831,0.2822,0.2807,0.2798,0.2797,0.2788,0.2778)
ldatDate<-c("1987-06-01","1989-11-30","1994-09-29","1998-05-31","2003-03-30","2007-03-30","2007-08-31","2010-11-30","2014-01-01")
ldatsizecomp<-c("6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE","6PIPE")

#Additional Data in separate Date Set
ldat<-data.frame(compid=ldatcompid, nps=ldatnps,       comp=ldatcomp,case=ldatcase,year=ldatyear,   Mean=ldatMean,Date=ldatDate,sizecomp=ldatsizecomp)
ldat$Date=as.Date(ldat$Date)

ggplot(data=dat,aes(x=as.Date(Date),y=thick, group=tpcolor,color=tpcolor)) + geom_point(shape=18,size=3,position = position_jitter(w = 50, h = 0)) + # Plot Main Data 
gtitle(paste0("MyCircuit","_","MyComponent","(",dat$compid,")")) +  #Plot Title
theme_bw() + theme(plot.title= element_text(size=16, face="bold")) + 
labs(x="Survey Dates and Beyond",y="Thickness(in Inches)") +  #Axis labels

***#MeanLine of 2nd Set Of Data  --- WHEN LINE 39 (geom_smooth) or LINE 40 (geom_line) below are UNCOMMENTED, the PLOT NO LONGER RECOGNIZES tpcolor used in line 31 (ggplot2) above.  Instead I receive an error Message:  Error in eval(expr, envir, enclos) : object 'tpcolor' not found --WHY?
#geom_smooth(data=ldat, aes(x=Date, y=Mean), method='lm',se=FALSE, fullrange=TRUE, colour="black", linetype="dotdash", size=0.25) +
#geom_line(data=ldat, aes(x=Date, y=Mean), colour="black", linetype="dotdash", size=0.25) +***

scale_colour_discrete(name  ="Test Points",
                    breaks=dat$tpcolor,
                    labels=dat$tp)

当我尝试使用第二个数据帧 ldat 中的数据添加到绘图中时，出现了问题。上面的代码包含后面的行，但已被注释掉。

geom_smooth(data=ldat, aes(x=Date, y=Mean), method='lm',se=FALSE, fullrange=TRUE, colour="black", linetype="dotdash", size=0.25) +

当此代码行取消注释时，我收到错误:

Error in eval(expr, envir, enclos) : object 'tpcolor' not found*

我不明白为什么。我尝试使用 geom_line 代替，但这似乎也不起作用。我尝试将其放在代码中的不同位置，但没有效果。

如果我从初始绘图(ggplot2 美学)中删除 , group=tpcolor,color=tpcolor ，错误就会消失，但我会丢失要分配的颜色。如果我使用 dat$tpcolor 显式分配 ggplot2 美学中的颜色，则会收到不同的错误:

Error: Aesthetics must either be length one, or the same length as the dataProblems:Date, Mean

在这里，我认为我只是将同一日期范围内的另一个图层添加到我的绘图中，与最初绘制的数据分开 - 但这似乎意味着分配给主要数据集的颜色正在应用于第二个数据集，我我不想这样做。

我不太确定接下来要尝试什么。有什么想法可以帮助我使用原始 dat 数据框中每个 tp 的指定颜色将两组数据放到同一个绘图上？

Answer 1

当您调用 ggplot 时，以下所有几何对象都会获得您提供的初始 aes 。这会引导您的 geom_smooth 调用查找 tpcolor ，以便它知道如何为组着色，但找不到它，因此您的错误。

要解决此问题，您可以尝试分别使用每个几何对象中的数据:

ggplot() + geom_point(data = dat, aes(x = as.Date(Date), y = thick, group = tpcolor, color = tpcolor), shape = 18, size = 3, position = position_jitter(w = 50, h = 0)) +
           geom_smooth(data = ldat, aes(x = Date, y = Mean), method = 'lm', se = FALSE, fullrange = TRUE, colour = "black", linetype = "dotdash", size = 0.25)

或者您可以将 inherit.aes=FALSE 添加到 geom_smooth 调用中，以确保它不会继承之前的 aes:

ggplot(data = dat, aes(x = as.Date(Date), y = thick, group = tpcolor, color = tpcolor)) +
       geom_point(shape = 18, size = 3, position = position_jitter(w = 50, h = 0)) +
       geom_smooth(data = ldat, aes(x = Date, y = Mean), method = 'lm', se = FALSE, fullrange = TRUE, colour = "black", linetype = "dotdash", size = 0.25, inherit.aes = FALSE)