如何使用各自的tukey测试对boxplot中的变量进行排序

Question

我使用了r-graph-gallery.com中的代码，并根据我的数据进行了调整。

我想在一个图中做6个盒子图，并以特定的顺序对它们进行排序，但是当我在进行排序时，tukey分析不会同时对其本身进行排序！

你知道我如何改进它吗？

这是错误的顺序，但是正确的tukey测试表示：

​

在这里，它有正确的顺序，但没有正确的tukey测试重新划分。

​

我应该怎么做才能得到TUKEY测试的正确顺序呢？

顺便问一下，谁知道如何得到tukey测试的最高均值的"a“值，而不是"c”？

感谢您的帮助！

下面是我使用的代码：

date<- (read.delim("SoilOBIoldtd.txt", header=TRUE))
# library
library(multcompView)
# What is the effect of the level on the CEC ?
model=lm( date$CEC_eff ~ date$level )

ANOVA=aov(model)
# Tukey test to study each pair of level :
TUKEY <- TukeyHSD(x=ANOVA, 'date$level', conf.level=0.95)
#This line is the difference between the two plots (using or ignoring this line)
date$level <- factor(date$level , levels=c("DAFS_Top", "DAFS_Down",
"CONV_Top","CONV_Down","Old_cocoa_Top","Old_cocoa_Down"))
# Tuckey test representation :
plot(TUKEY , las=1 , col="brown")
generate_label_df <- function(TUKEY, CEC_eff){
   # Extract labels and factor levels from Tukey post-hoc 
  Tukey.levels <- TUKEY[[CEC_eff]][,4]
  Tukey.labels <- data.frame(multcompLetters(Tukey.levels,reversed = FALSE)['Letters'])
  #I need to put the labels in the same order as in the boxplot :
  Tukey.labels$level=rownames(Tukey.labels)
  Tukey.labels=Tukey.labels[order(Tukey.labels$level) , ]
  return(Tukey.labels)}
# Apply the function on my dataset
LABELS <- generate_label_df(TUKEY , "date$level")
# A panel of colors to draw each group with the same color :
my_colors <- c(   rgb(143,199,74,maxColorValue = 255),  rgb(242,104,34,maxColorValue = 255),   rgb(111,145,202,maxColorValue = 255))
# Draw the basic boxplot
a <-boxplot(date$CEC_eff ~ date$level , ylim=c(min(date$CEC_eff ) , 1.1*max(date$CEC_eff)) , col=my_colors[as.numeric(LABELS[,1])] , ylab="CEC" , main="")
# I want to write the letter over each box. Over is how high I want to write it.
over <- 0.1*max( a$stats[nrow(a$stats),] )
#Add the labels
text( c(1:nlevels(date$level)) , a$stats[nrow(a$stats),]+over , LABELS[,1]  , col=my_colors[as.numeric(LABELS[,1])] )

Answer 1

嘿，我想我知道你的问题了，基本上是正确地给盒子图分配了一个颜色的/text标签。

下面我生成了一些数据，做了Tukey，并得到了标签：

library(multcompView)

set.seed(111)
lvl = c("DAFS_Top", "DAFS_Down",
"CONV_Top","CONV_Down","Old_cocoa_Top","Old_cocoa_Down")
df = data.frame(CEC_eff=rnorm(48,rep(c(2,4,6),each=8),1),level=rep(lvl,each=8))

df$level <- factor(df$level , levels=c("DAFS_Top", "DAFS_Down",
"CONV_Top","CONV_Down","Old_cocoa_Top","Old_cocoa_Down"))

ANOVA=aov(lm(CEC_eff ~ level ,data=df))
TUKEY <- TukeyHSD(x=ANOVA, "level", conf.level=0.95)
LABELS <- multcompLetters(TUKEY$level[,4],reversed=FALSE)$Letters

现在，您定义了颜色，并且可以将字母分配给颜色。

my_colors <- c(   rgb(143,199,74,maxColorValue = 255),  
rgb(242,104,34,maxColorValue = 255),   rgb(111,145,202,maxColorValue = 255))

names(my_colors) <- sort(unique(LABELS))

然后我们将颜色分配给将出现在箱线图中的级别：

lvl_colors <- my_colors[LABELS[levels(df$level)]]

和plot：

a <-boxplot(CEC_eff ~ level , data=df, cex.axis=0.7,ylim=c(min(df$CEC_eff ),
1.1*max(df$CEC_eff)) , col= lvl_colors, ylab="CEC" , main="")

over <- 0.1*max( a$stats[nrow(a$stats),] )
#Add the labels
text(1:nlevels(df$level) ,a$stats[nrow(a$stats),]+over ,names(lvl_colors )  , 
col=lvl_colors  )

​

​