使用ggplot向Barplot添加错误栏

Question

我正在为我所做的研究绘制多样性指数均值的柱状图。我已经计算出每个样本的索引，并将其添加到我的表中（我通常会读入该表）。然后，我计算了两种不同环境的指数均值并绘制了这些均值。但是，我无法解决如何添加错误栏的问题。我了解ggplot2是执行此操作的有用工具，但是无法理解我的解释。

SO，基本上是尝试从表中取两个值的平均值，并绘制带有误差线的条形图。这是我目前的代码

河流和湖泊的平均香农

`mean_river <- mean(parasite_data$Shannon.index[1:24])
mean_lake <- mean(parasite_data$Shannon.index[25:43])`

均值矩阵

Shannon_mean <- matrix(c(mean_river, mean_lake), nrow = 1, ncol = 2, dimnames = list(c("mean"), c("River","Lake")))

绘图图＃

`barplot(Shannon_mean, 

# name axis 
    xlab = "Environment", ylab = "Shannon Diversity Index", 

# title of graph 
    main = "Diversity of Parasites found on Fish from River 
            and Lake Environments", 

# size of title text and colour of bars #
    cex.main = 1, col = "gray80")`

就像我说的那样，

我曾尝试使用ggplot，但无法正确读取数据。任何帮助，将不胜感激。

Answer 1

使用ggplot2和标准差创建误差线：

library(ggplot2)

# I'm just making up the numbers to provide the solution
mean_river <- 25 # mean(parasite_data$Shannon.index[1:24])
mean_lake <- 29 # mean(parasite_data$Shannon.index[25:43])

# To plot the error bars i'm assuming you want the standard deviation
# but you can use the min or max value
sd_river <- 0.5 # sd(parasite_data$Shannon.index[1:24])
sd_lake <- 4 # sd(parasite_data$Shannon.index[25:43])

# Instead of matrix I would use a data.frame for ggplot
Shannon_data <- data.frame(name = c("River", "Lake"),
                           mean = c(mean_river, mean_lake),
                           sd = c(sd_river, sd_lake)) 

ggplot(Shannon_data) +
  geom_bar(aes(x=name, y=mean), stat="identity", fill="gray80") +
  geom_errorbar(aes(x=name, ymin=mean-sd, ymax=mean+sd), 
                width=0.4, colour="orange", alpha=0.9, size=1.3) +
  labs(title = "Diversity of Parasites found on Fish from River and Lake Environments") + 
  xlab("Environment") + ylab("Shannon Diversity Index") + theme_bw()

这将生成下一个图：

使用最小值和最大值创建误差线：

# Example with max and min value instead of sd
min_river <- 22 # min(parasite_data$Shannon.index[1:24])
min_lake <- 21 # min(parasite_data$Shannon.index[25:43])
max_river <- 31 # max(parasite_data$Shannon.index[1:24])
max_lake <- 30 # max(parasite_data$Shannon.index[25:43])

Shannon_data <- data.frame(name = c("River", "Lake"),
                           mean = c(mean_river, mean_lake),
                           min = c(min_river, min_lake), 
                           max = c(max_river, max_lake)) 
ggplot(Shannon_data) +
  geom_bar(aes(x=name, y=mean), stat="identity", fill="gray80") +
  geom_errorbar(aes(x=name, ymin=min, ymax=max), 
                 width=0.4, colour="orange", alpha=0.9, size=1.3) +
  labs(title = "Diversity of Parasites found on Fish from River and Lake Environments") + 
  xlab("Environment") + ylab("Shannon Diversity Index") + theme_bw()

希望获得帮助！

Answer 2

我模拟了一些看起来像您的数据，因此您无需将它们放入单独的向量中。将它们保存在数据框中

Shannon.index <- runif(43,1.5,3.5)
type = rep(c("River","Lake"),times=c(24,19)) 

df <- data.frame(Shannon.index,type)

对于基于R的条形图，我们需要计算均值（sem）和均值的标准误差（就像您所做的那样），并使用arrows()添加误差线：

Shannon_sem <- tapply(df$Shannon.index,df$type,function(x)sd(x)/sqrt(length(x)))
Shannon_mean <- tapply(df$Shannon.index,df$type,mean)
YMAX <-ceiling(max(Shannon_mean+Shannon_sem))

PLOT <- barplot(Shannon_mean,
xlab = "Environment", ylab = "Shannon Diversity Index", 
main = "Diversity of Parasites \nfound on Fish from River 
and Lake Environments", cex.main = 1, col = "gray80",
ylim = c(0,YMAX))
arrows(x0=PLOT,y0=Shannon_mean+Shannon_sem,cex.main=0.7,
y1=Shannon_mean-Shannon_sem,angle=90,code=3,length=0.1)

如果您使用ggplot2：

library(ggplot2)
ggplot(df,aes(x=type,y=Shannon.index)) + stat_summary(fun.y=mean,geom="bar",fill="gray80") + 
theme_bw() + 
stat_summary(fun.data = mean_se, geom = "errorbar",width=0.2)

您可以使用stat_summary()

即时计算平均值和标准误差。