r - 堆积条形图 ggplot2-按特定变量从最高值到最低值对样本进行排序

Question

我正在尝试以堆叠条形图格式绘制包含超过 2000 个样本的数据集，其中每个样本(由“SampleID”表示)位于 x 轴上，y 轴上有 6 个测量值(测量1-6)。我希望按以下测量变量顺序显示/排序样本:Measurement4、1、5、2、3 和 6，以及从​​最高到最低测量值。下面是 15 个样本的子集，作为我正在使用的示例，我将其称为“dummy_set”数据框:

    SampleID Measurement1 Measurement2 Measurement3 Measurement4 Measurement5 Measurement6
1         A         0.05         0.00         0.95         0.00          0.0         0.00
2         B         0.00         0.00         0.43         0.56          0.0         0.01
3         C         0.64         0.36         0.00         0.00          0.0         0.00
4         D         0.00         0.82         0.18         0.00          0.0         0.00
5         E         0.00         0.60         0.00         0.40          0.0         0.00
6         F         0.80         0.00         0.00         0.20          0.0         0.00
7         G         0.00         0.00         0.00         1.00          0.0         0.00
8         H         0.00         0.00         0.00         1.00          0.0         0.00
9         I         0.00         0.00         1.00         0.00          0.0         0.00
10        J         0.00         0.00         1.00         0.00          0.0         0.00
11        K         0.25         0.00         0.00         0.45          0.3         0.00
12        L         0.10         0.00         0.00         0.10          0.8         0.00
13        M         0.19         0.10         0.00         0.70          0.0         0.01
14        N         0.90         0.00         0.00         0.10          0.0         0.00
15        O         0.00         0.10         0.40         0.00          0.5         0.00

这是我所做的基础知识:

1. 融合数据集:Melt_dummy_set <- Melt(dummy_set, id.var = "SampleID")

   融化的数据集如下所示:

   head(melt_dummy_set)
       SampleID     variable value
   1         A Measurement1  0.05
   2         B Measurement1  0.00
   3         C Measurement1  0.64
   4         D Measurement1  0.00
   5         E Measurement1  0.00
   6         F Measurement1  0.80
   

2. 使用 ggplot() 和 geom_bar() 绘制融化的数据集:

   ggplot(melt_dummy_set, aes(x = SampleID, y = value, fill = variable)) + 
   geom_bar(stat = "identity") + 
   

如您所见，样本按其列出的原始顺序 (A-O) 绘制。但是，我希望它们按以下顺序绘制:G、H、M、B、K、N、F、C、L、O、D、E、I、J 和 A。

根据其他类似的 Stack Overflow 问题，我发现我需要按照我想要的顺序重新调整/重新建立因素。这是我到目前为止所尝试过的:

#Attempt 1
reordered_melt_dummy_set <- transform(melt_dummy_set, variable = reorder(variable, -value))
ggplot(reordered_melt_dummy_set, aes(x = SampleID, y = value, fill = variable)) + 
  geom_bar(stat = "identity") + 

Attempt 1 results

#Attempt 2
copy_melt_dummy_set <- melt_dummy_set
copy_melt_dummy_set$variable <- factor(copy_melt_dummy_set$variable, levels = c("Measurement4", "Measurement5", "Measurement1", "Measurement2", "Measurement3", "Measurement6"))
ggplot(copy_melt_dummy_set, aes(x = SampleID, y = value, fill = variable)) + 
  geom_bar(stat = "identity") + 

Attempt 2 results

我的第三次尝试导致了多个错误(在代码行后面以“##”表示)

#Attempt 3
copy2_melt_dummy_set <- melt_dummy_set

copy2_melt_dummy_set$SampleID <- factor(copy2_melt_dummy_set$SampleID, levels = copy2_melt_dummy_set[order(-copy2_melt_dummy_set$value), "variable"])
##Error in `levels<-`(`*tmp*`, value = as.character(levels)) : factor level [2] is duplicated

copy2_melt_dummy_set$variable <- factor(copy2_melt_dummy_set$variable, levels = copy2_melt_dummy_set[order(copy2_melt_dummy_set$value), "variable"])
## Error in `levels<-`(`*tmp*`, value = as.character(levels)) : factor level [2] is duplicated

copy2_melt_dummy_set$SampleID <- factor(copy2_melt_dummy_set$SampleID, levels = copy2_melt_dummy_set[order(-copy2_melt_dummy_set$variable), "SampleID"])
## Error in `levels<-`(`*tmp*`, value = as.character(levels)) : factor level [16] is duplicated
## In addition: Warning message: In Ops.factor(copy2_melt_dummy_set$variable) : ‘-’ not meaningful for factors

copy2_melt_dummy_set$SampleID <- factor(copy2_melt_dummy_set$SampleID, levels = copy2_melt_dummy_set[order(-copy2_melt_dummy_set$value), "SampleID"])
## Error in `levels<-`(`*tmp*`, value = as.character(levels)) : factor level [16] is duplicated

copy2_melt_dummy_set$SampleID <- factor(copy2_melt_dummy_set$SampleID, levels = copy2_melt_dummy_set[order(-copy2_melt_dummy_set$value), "value"])
## Error in `levels<-`(`*tmp*`, value = as.character(levels)) : factor level [2] is duplicated

#Attempt 4
copy3_melt_dummy_set <- melt_dummy_set[order(melt_dummy_set$variable, -melt_dummy_set$value), ]
head(copy3_melt_dummy_set)
ggplot(copy3_melt_dummy_set, aes(x = SampleID, y = value, fill = variable)) + 
  geom_bar(stat = "identity") + 

Attempt 4 results

#Attempt 5
ggplot(melt_dummy_set[order(melt_dummy_set$variable, -melt_dummy_set$value), ], aes(x = SampleID, y = value, fill = variable)) + 
  geom_bar(stat = "identity") + 

Attempt 5 results

#Attempt 6
new_melt_dummy_set <- within(melt_dummy_set, 
                             variable <- factor(variable, levels = names(sort(table(variable), decreasing = TRUE))))
ggplot(new_melt_dummy_set, aes(x = SampleID, y = value, fill = variable)) + 
  geom_bar(stat = "identity") + 

Attempt 6 results

#Attempt 7
copy4_melt_dummy_set <- melt_dummy_set
custom_leveling <- unique(copy4_melt_dummy_set$variable)
copy4_melt_dummy_set$variable <- factor(copy4_melt_dummy_set$variable, level = custom_leveling)
ggplot(copy4_melt_dummy_set, aes(x = SampleID, y = value, fill = variable)) + 
  geom_bar(stat = "identity") + 

Attempt 7 results

在所有情况下，我都无法重新组织 x 轴上的实际样本。我觉得可能有一个简单的解决办法，但我不知道我做错了什么。有什么建议吗？

已编辑

为了回应可能的重复评论，我尝试应用 Order Bars in ggplot2 bar graph 中的代码/解决方案他们没有按照我想要的顺序制作情节。请参阅下面的我尝试过的代码:

#First solution
new_melt_dummy_set <- within(melt_dummy_set, 
                             variable <- factor(variable, levels = names(sort(table(variable), decreasing = TRUE))))
ggplot(new_melt_dummy_set, aes(x = SampleID, y = value, fill = variable)) + 
  geom_bar(stat = "identity")

#Second solution
ggplot(melt_dummy_set, aes(x = reorder(SampleID, variable, function(x)-length(x)), y = value, fill = variable)) + geom_bar(stat = "identity")
ggplot(melt_dummy_set, aes(x = reorder(variable, SampleID, function(x)-length(x)), y = value, fill = variable)) + geom_bar(stat = "identity")

#Third solution
ordered_measurements <- c("Measurement4", "Measurement1", "Measurement5", "Measurement2", "Measurement3", "Measurement6")
ggplot(melt_dummy_set, aes(x = SampleID, y = value, fill = variable)) + 
  geom_bar(stat = "identity") + 
  scale_x_discrete(limits = ordered_measurements)

#Fourth solution
ggplot(melt_dummy_set, aes(x = reorder(SampleID, -table(variable)[variable]), y = value, fill = variable)) + geom_bar(stat = "identity")

require(forcats)
ggplot(melt_dummy_set, aes(x = SampleID, fill = fct_infreq(variable), y = value)) + geom_bar(stat = "identity")
ggplot(melt_dummy_set, aes(x = fct_infreq(variable))) + geom_bar(stat = "identity")

#Fifth solution
library(tidyverse)
library(forcats)
melt_dummy_set %>%
  mutate(variable = fct_reorder(variable, value, .desc = TRUE)) %>%
  ggplot(aes(x = SampleID, y = value, fill = variable)) + geom_bar(stat = 'identity')

#Sixth solution
library(dplyr)
melt_dummy_set %>%
  group_by(variable) %>%                              
  summarize(counts = n()) %>%
  arrange(-counts) %>%                                
  mutate(SampleID = factor(SampleID, variable)) %>%   
  ggplot(aes(x = SampleID, y = value, fill = variable)) +                  
  geom_bar(stat = "identity")                         

melt_dummy_set %>%
  group_by(SampleID) %>%                              
  summarize(counts = n()) %>%
  arrange(-counts) %>%                                
  mutate(SampleID = factor(SampleID, value)) %>%   
  ggplot(aes(x = SampleID, y = value, fill = variable)) +                 
  geom_bar(stat = "identity")

#Seventh solution
new_meltedDummy_set <- transform(melt_dummy_set,
                       variable = ordered(variable, levels = names(sort(-table(variable)))))
ggplot(new_meltedDummy_set, aes(x = SampleID, y = value, fill = variable)) +
  geom_bar(stat = "identity")

Answer 1

这就是你的目的吗？我想你很接近。您不需要将 Measurement 变量列转换为因子，而是需要根据 Measurement 值的顺序对 SampleID 列进行排序。这是计算 sample_order 的行中发生的情况:

library(tidyverse)

dummy_set <- tibble(
  SampleID = c("A", "B", "C", "D", "E", "F", "G", "H", "I", "J", "K", "L", "M", "N", "O"),
  Measurement1 = c(0.05, 0, 0.64, 0, 0, 0.8, 0, 0, 0, 0, 0.25, 0.1, 0.19, 0.9, 0),
  Measurement2 = c(0, 0, 0.36, 0.82, 0.6, 0, 0, 0, 0, 0, 0, 0, 0.1, 0, 0.1),
  Measurement3 = c(0.95, 0.43, 0, 0.18, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0.4),
  Measurement4 = c(0, 0.56, 0, 0, 0.4, 0.2, 1, 1, 0, 0, 0.45, 0.1, 0.7, 0.1, 0),
  Measurement5 = c(0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.3, 0.8, 0, 0, 0.5),
  Measurement6 = c(0, 0.01, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.01, 0, 0)
)

sample_order <- dummy_set %>%
  arrange(desc(Measurement4), desc(Measurement1), desc(Measurement5), desc(Measurement2), desc(Measurement3), desc(Measurement6)) %>%
  pull(SampleID)

melt_dummy_set <- dummy_set %>%
  gather(variable, value, -SampleID)

reordered_melt_dummy_set <- melt_dummy_set %>%
  mutate(SampleID = factor(SampleID, levels = sample_order))

plot_ordered <- ggplot(reordered_melt_dummy_set, aes(x = SampleID, y = value, fill = variable)) +
  geom_bar(stat = "identity") +
  scale_y_continuous(expand = c(0,0)) +
  theme(axis.ticks.x = element_blank(), panel.grid = element_blank(), axis.line = element_line(color = "black"), panel.border = element_blank(), panel.background = element_blank())

plot_ordered

^{由reprex package于2019年7月26日创建(v0.3.0)}