如何用R分析多响应名义与分类变量并绘制堆叠条形图

问题说明 #

我是R编程新手，尝试分析两组数据并生成堆叠条形图，但一直没成功。

原始数据 #

第一组：食物熟悉度数据 #

列名：Cookies、Pizza、Strawberry
共6行数据：

• Very Familiar; Familiar; Not Familiar
• Familiar; Very Familiar; Not Familiar
• Familiar; Not Familiar; Very Familiar
• Very Familiar; Familiar; Not Familiar
• Familiar; Very Familiar; Not Familiar
• Familiar; Not Familiar; Very Familiar

第二组：员工绩效提升因素数据 #

列名：training program、flexible work、communication
共7行数据（已统一态度选项大小写）：

• Neutral; Agree; Agree
• Disagree; Agree; Agree
• Neutral; Disagree; Agree
• Agree; Disagree; Disagree
• Disagree; Agree; Agree
• Neutral; Disagree; Agree
• Agree; Disagree; Disagree

你的错误代码 #

rf<-r %>% select(Cookies,Pizza,Strawberry) %>% group_by (Cookies,Pizza,Strawberry)
rfp <- ggplot(data=rf, aes(x=Cookies,Pizza,Strawberry))

问题分析 #

你的代码有两个核心问题：

1. group_by(Cookies,Pizza,Strawberry)是把三个列的组合作为分组，这不是堆叠条形图需要的逻辑——堆叠图需要统计每个类别下各选项的计数，应该对每一列单独统计频率，再把数据转为长格式（long format）。
2. ggplot的aes()参数里不能同时放多个x轴变量，堆叠条形图的逻辑是：x轴是类别（比如食物名称/影响因素），y轴是计数，fill是熟悉度/态度选项。

正确实现步骤 #

第一组数据：食物熟悉度堆叠条形图 #

1. 先构造原始数据框：

# 构造熟悉度数据
familiarity_data <- data.frame(
  Cookies = c("Very Familiar", "Familiar", "Familiar", "Very Familiar", "Familiar", "Familiar"),
  Pizza = c("Familiar", "Very Familiar", "Not Familiar", "Familiar", "Very Familiar", "Not Familiar"),
  Strawberry = c("Not Familiar", "Not Familiar", "Very Familiar", "Not Familiar", "Not Familiar", "Very Familiar")
)

2. 把宽格式数据转为长格式（ggplot对长格式数据更友好）：

library(tidyverse)

familiarity_long <- familiarity_data %>%
  pivot_longer(cols = everything(), names_to = "Food", values_to = "Familiarity")

3. 统计每个食物下各熟悉度的人数：

familiarity_count <- familiarity_long %>%
  count(Food, Familiarity)

4. 绘制堆叠条形图：

ggplot(familiarity_count, aes(x = Food, y = n, fill = Familiarity)) +
  geom_bar(stat = "identity") +
  labs(title = "食物熟悉度分布", x = "食物", y = "人数", fill = "熟悉度") +
  theme_minimal()

第二组数据：员工绩效提升因素堆叠条形图 #

1. 构造原始数据框（统一态度选项的大小写，避免统计错误）：

# 构造绩效提升数据
performance_data <- data.frame(
  `training program` = c("Neutral", "Disagree", "Neutral", "Agree", "Disagree", "Neutral", "Agree"),
  `flexible work` = c("Agree", "Agree", "Disagree", "Disagree", "Agree", "Disagree", "Disagree"),
  communication = c("Agree", "Agree", "Agree", "Disagree", "Agree", "Agree", "Disagree")
)

2. 转为长格式：

performance_long <- performance_data %>%
  pivot_longer(cols = everything(), names_to = "Factor", values_to = "Attitude")

3. 统计每个因素下各态度的人数：

performance_count <- performance_long %>%
  count(Factor, Attitude)

4. 绘制堆叠条形图：

ggplot(performance_count, aes(x = Factor, y = n, fill = Attitude)) +
  geom_bar(stat = "identity") +
  labs(title = "员工绩效提升因素态度分布", x = "影响因素", y = "人数", fill = "态度") +
  theme_minimal()

内容的提问来源于stack exchange，提问作者Opeyemi Okolie