如何修复Plotly按钮切换多幅ggplot转换图表的显示异常问题？

问题根源 #

你当前的代码存在两个核心问题：

• 笔误：第一个ggplot对象命名为barp，但后续转换为plotly时调用了不存在的plot1，导致第一个plotly对象无效。
• 逻辑错误：restyle方法用于修改当前figure内已有trace的属性，而非切换独立图表。你仅基于第一个plot构建figure，按钮操作的是该plot内部的trace，而非切换到其他完整图表。

修正方案 #

正确思路是：将4个图表的所有trace合并到同一个Plotly figure中，通过按钮控制对应图表的所有trace可见性，其余隐藏。

修正后的完整代码

# 修正第一个ggplot的命名，确保后续调用一致
plot1 <- ggplot(data = dat, aes(x = AnnualIncome_Label, fill = interest_Label)) +
  geom_bar(position = "fill") +
  xlab("Annual Income") +
  ylab("Proportion") +
  ggtitle("Proportional Interest in Clinical Trials by Income") +
  labs(fill = "Clinical Trial Interest Level") +
  theme_minimal() +
  theme(axis.text.x = element_text(angle = 45, hjust = 1)) +
  scale_fill_manual(values = c("khaki2", "lemonchiffon1", "#DADAEB", "#9E9AC8", "#6A51A3"))

# Stacked bar chart for "Education" and "Interest_Label"
plot2 <- ggplot(data = dat, aes(x = Ed_Label, fill = interest_Label)) +
  geom_bar(position = "fill") +
  xlab("Education Level") +
  ylab("Proportion") +
  ggtitle("Proportional Interest in Clinical Trials by Education Level") +
  labs(fill = "Clinical Trial Interest Level") +
  theme_minimal() +
  theme(axis.text.x = element_text(angle = 45, hjust = 1)) +
  scale_fill_manual(values = c("khaki2", "lemonchiffon1", "#DADAEB", "#9E9AC8", "#6A51A3"))

# Stacked bar chart for "EnglishProf_Label" and "Interest_Label"
plot3 <- ggplot(data = dat, aes(x = EnglishProf_Label, fill = interest_Label)) +
  geom_bar(position = "fill") +
  xlab("English Proficiency Label") +
  ylab("Proportion") +
  ggtitle("Proportional Interest in Clinical Trials by English Proficiency") +
  labs(fill = "Clinical Trial Interest Level") +
  theme_minimal() +
  theme(axis.text.x = element_text(angle = 45, hjust = 1)) +
  scale_fill_manual(values = c("khaki2", "lemonchiffon1", "#DADAEB", "#9E9AC8", "#6A51A3"))

# Stacked bar chart for "HelpMedLit_Label" and "Interest_Label"
plot4 <- ggplot(data = dat, aes(x = HelpMedLit_Label, fill = interest_Label)) +
  geom_bar(position = "fill") +
  xlab("Help with Medical Literacy Label") +
  ylab("Proportion") +
  ggtitle("Proportional Interest in Clinical Trials by Help Understanding Medical Literature") +
  labs(fill = "Clinical Trial Interest Level") +
  theme_minimal() +
  theme(axis.text.x = element_text(angle = 45, hjust = 1), 
        legend.position = "none") +
  guides(fill = FALSE) +
  scale_fill_manual(values = c("khaki2", "lemonchiffon1", "#DADAEB", "#9E9AC8", "#6A51A3"))

# 转换为plotly对象
plotly_plot1 <- ggplotly(plot1)
plotly_plot2 <- ggplotly(plot2)
plotly_plot3 <- ggplotly(plot3)
plotly_plot4 <- ggplotly(plot4)

# 创建空的Plotly figure
fig <- plot_ly()

# 将所有plot的trace添加到同一个figure中
fig <- fig %>% 
  add_traces(plotly_plot1$x$data) %>%
  add_traces(plotly_plot2$x$data) %>%
  add_traces(plotly_plot3$x$data) %>%
  add_traces(plotly_plot4$x$data)

# 获取每个堆叠图的trace数量（等于interest_Label的类别数）
trace_count_per_plot <- length(plotly_plot1$x$data)

# 生成按钮对应的可见性向量
create_visible_vec <- function(plot_index, total_plots, trace_per_plot) {
  vec <- rep(FALSE, total_plots * trace_per_plot)
  start <- (plot_index - 1) * trace_per_plot + 1
  end <- plot_index * trace_per_plot
  vec[start:end] <- TRUE
  vec
}

# 创建下拉按钮列表
buttons <- list(
  list(method = "restyle",
       args = list("visible", create_visible_vec(1, 4, trace_count_per_plot)),
       label = "Income vs Interest"),
  list(method = "restyle",
       args = list("visible", create_visible_vec(2, 4, trace_count_per_plot)),
       label = "Education vs Interest"),
  list(method = "restyle",
       args = list("visible", create_visible_vec(3, 4, trace_count_per_plot)),
       label = "English Proficiency vs Interest"),
  list(method = "restyle",
       args = list("visible", create_visible_vec(4, 4, trace_count_per_plot)),
       label = "Medical Literacy Help vs Interest")
)

# 设置布局并应用初始可见性
fig <- fig %>% 
  layout(
    updatemenus = list(
      list(
        buttons = buttons,
        x = 0.1,
        y = 1.15,
        xanchor = "left",
        yanchor = "top"
      )
    ),
    showlegend = TRUE
  ) %>%
  restyle("visible", create_visible_vec(1, 4, trace_count_per_plot))

fig

关键说明 #

1. 修正笔误：统一第一个ggplot对象的命名为plot1，确保后续plotly转换引用正确。
2. 合并trace：通过add_traces将4个图表的所有trace整合到同一个figure，让所有图表元素处于同一Plotly对象中。
3. 精准控制可见性：自定义函数生成每个按钮对应的可见性向量，确保切换时仅显示目标图表的全部trace，其余隐藏。
4. 布局优化：调整按钮位置，避免遮挡图表内容。

内容的提问来源于stack exchange，提问作者Mel