优化R语言两分类变量计数绘图流程：简化ggplot2实现

你可以完全跳过手动创建数据框的步骤，用以下几种更简便的方式实现需求：

方法1：直接用ggplot的stat_count自动统计（最简便） #

不用提前计算任何频数，直接基于原始数据绘图，ggplot会自动帮你统计每组的人数：

suppressPackageStartupMessages(library(ggplot2))

# 原始数据
low_income <- c("yes","yes", "no","yes","yes","yes", "no","yes","yes","yes", "no","no")
test <- c("completed", "none","completed", "none","completed", "completed","completed", "completed", "none","completed", "none","completed")
df <- data.frame(low_income, test)

# 直接绘图，stat="count"是geom_bar的默认值，可省略
ggplot(df, aes(x = low_income, fill = test)) + 
  geom_bar(position = "dodge") +  # 默认stat=count，自动统计频数
  scale_fill_brewer(palette = "Set1")

如果只想展示低收入（low_income == "yes"）的群体，加个筛选即可：

ggplot(df[df$low_income == "yes", ], aes(x = low_income, fill = test)) + 
  geom_bar(position = "dodge") +
  scale_fill_brewer(palette = "Set1")

方法2：用dplyr自动生成统计数据框 #

如果需要保留统计后的数据集，用dplyr的分组汇总功能，一键生成包含count的dataframe，无需手动输入数值：

suppressPackageStartupMessages(library(ggplot2))
suppressPackageStartupMessages(library(dplyr))

# 原始数据
low_income <- c("yes","yes", "no","yes","yes","yes", "no","yes","yes","yes", "no","no")
test <- c("completed", "none","completed", "none","completed", "completed","completed", "completed", "none","completed", "none","completed")
df <- data.frame(low_income, test)

# 自动分组统计
df_2 <- df %>%
  group_by(low_income, test) %>%
  summarise(count = n(), .groups = "drop")  # .groups="drop"取消分组状态

# 绘图
ggplot(df_2, aes(x = low_income, y = count, fill = test)) + 
  geom_bar(stat = "identity", position = "dodge") + 
  scale_fill_brewer(palette = "Set1")

方法3：将table对象直接转为数据框 #

如果你已经生成了table，不用手动输入数值，直接把table转成dataframe即可：

suppressPackageStartupMessages(library(ggplot2))

# 原始数据
low_income <- c("yes","yes", "no","yes","yes","yes", "no","yes","yes","yes", "no","no")
test <- c("completed", "none","completed", "none","completed", "completed","completed", "completed", "none","completed", "none","completed")
df <- data.frame(low_income, test)

# 生成table后直接转成数据框
table1 <- table(df$low_income, df$test)
df_2 <- as.data.frame(table1)
colnames(df_2) <- c("low_income", "test", "count")  # 重命名列名

# 绘图
ggplot(df_2, aes(x = low_income, y = count, fill = test)) + 
  geom_bar(stat = "identity", position = "dodge") + 
  scale_fill_brewer(palette = "Set1")

内容的提问来源于stack exchange，提问作者Rebecca James