如何缩短复杂ggplot图形的编译时间？

针对ggpattern的优化方案 #

ggpattern渲染图案时计算量较大，尤其是数据集规模大或默认参数设置过高时，会导致绘图卡顿。可以通过调整以下参数降低计算负载：

• 降低图案分辨率：设置pattern_res = 20（默认值通常为100，调低后能大幅减少渲染计算量）
• 关闭抗锯齿：添加pattern_antialias = FALSE
• 减少图案密度/间距：适当调小pattern_density或调大pattern_spacing

优化后的示例代码：

library(ggplot2)
library(ggpattern)

# 示例数据集
df <- data.frame(
  category = rep(c("A", "B", "C"), each = 5),
  value = rnorm(15, 50, 10)
)

ggplot(df, aes(x = category, y = value, fill = category, pattern = category)) +
  geom_col_pattern(
    pattern_fill = "black",
    pattern_density = 0.2,
    pattern_spacing = 0.05,
    pattern_res = 20,          # 降低图案分辨率
    pattern_antialias = FALSE  # 关闭抗锯齿
  ) +
  scale_pattern_manual(values = c("stripe", "crosshatch", "circle")) +
  scale_fill_manual(values = c("#FF7F50", "#90EE90", "#87CEFA"))

替代方案：无需ggpattern实现带图案的柱状图 #

如果优化后仍卡顿，可以用ggplot2原生图层叠加实现图案效果，完全避免ggpattern的计算开销：

library(ggplot2)

# 先对数据做聚合（如果是分组柱状图，按需调整聚合逻辑）
df_summary <- aggregate(value ~ category, df, mean)

ggplot(df_summary, aes(x = category, y = value)) +
  # 底层纯色柱子
  geom_col(aes(fill = category), width = 0.7) +
  # 给类别B添加斜线
  geom_segment(
    data = subset(df_summary, category == "B"),
    aes(x = as.numeric(category) - 0.35, xend = as.numeric(category) + 0.35,
        y = seq(0, value, length.out = 10), yend = seq(0, value, length.out = 10)),
    color = "black", linewidth = 0.4, inherit.aes = FALSE
  ) +
  # 给类别C添加交叉线
  geom_segment(
    data = subset(df_summary, category == "C"),
    aes(x = as.numeric(category) - 0.35, xend = as.numeric(category) + 0.35,
        y = seq(0, value, length.out = 10), yend = seq(0, value, length.out = 10)),
    color = "black", linewidth = 0.4, inherit.aes = FALSE
  ) +
  geom_segment(
    data = subset(df_summary, category == "C"),
    aes(x = as.numeric(category) - 0.35, xend = as.numeric(category) + 0.35,
        y = seq(value, 0, length.out = 10), yend = seq(0, value, length.out = 10)),
    color = "black", linewidth = 0.4, inherit.aes = FALSE
  ) +
  scale_fill_manual(values = c("#FF7F50", "#90EE90", "#87CEFA")) +
  theme_minimal()

内存优化建议 #

• 关闭RStudio中不必要的标签页，清理冗余对象：运行rm(list = ls())后用gc()强制回收内存
• 重启R会话，释放被占用的内存后再运行绘图代码
• 如果数据集过大，先对数据做聚合（如计算均值、求和），避免绘制过多柱子

内容的提问来源于stack exchange，提问作者David Moldes