如何在R中合并多个corrplot相关图？含问题复现与导出需求

corrplot()属于基础R绘图系统，它的返回值是输入的相关系数矩阵，并非ggplot2类型的图形对象。而cowplot和patchwork默认仅支持处理ggplot2对象，直接传入corrplot的返回值自然会出现空白图或报错。

我们可以通过捕获基础绘图或转换为grob对象的方式，让cowplot/patchwork能够识别并合并这些相关图。

方法1：用recordPlot()捕获图形 #

# 生成并捕获mtcars相关图
cor_mtcars <- cor(mtcars[,1:4], mtcars[,5:8], method = "spearman", use="pairwise.complete.obs")
corrplot(cor_mtcars,
         method="color",
         addCoef.col = "black",
         number.cex=0.6,
         tl.cex = 0.8,
         addgrid.col = "grey",
         cl.cex=0.6,
         title = "Spearman Correlation Coefficients mtcars1",
         cex.main=0.8,
         mar=c(0,0,2,0))
p_mtcars <- recordPlot()  # 捕获当前绘图
dev.off()  # 关闭绘图设备

# 生成并捕获iris相关图
cor_iris <- cor(iris[,1:2], iris[,3:4], method = "spearman", use="pairwise.complete.obs")
corrplot(cor_iris,
         method="color",
         addCoef.col = "black",
         number.cex=0.6,
         tl.cex = 0.8,
         addgrid.col = "grey",
         cl.cex=0.6,
         title = "Spearman Correlation Coefficients iris",
         cex.main=0.8,
         mar=c(0,0,2,0))
p_iris <- recordPlot()
dev.off()

# 用cowplot合并（需用wrap_elements包裹捕获的图形）
library(cowplot)
# 17个图的话，将所有捕获的图形放入list，设置ncol=4即可自动排版
p_list <- list(p_mtcars, p_iris)  # 实际替换为17个捕获的图形对象
p_combined <- plot_grid(
  plotlist = p_list,
  ncol = 4,
  labels = "AUTO"
)
p_combined

# 用patchwork合并
library(patchwork)
p_combined_patch <- wrap_plots(p_list, ncol = 4)
p_combined_patch

方法2：用gridGraphics转成grob对象 #

library(gridGraphics)
library(cowplot)

# 定义转换函数
corrplot_to_grob <- function(corr_matrix, ...) {
  grid.echo({
    corrplot(corr_matrix, ...)
  })
  grid.grab()
}

# 生成grob格式的图形对象
p_mtcars_grob <- corrplot_to_grob(cor_mtcars,
                                  method="color",
                                  addCoef.col = "black",
                                  number.cex=0.6,
                                  tl.cex = 0.8,
                                  addgrid.col = "grey",
                                  cl.cex=0.6,
                                  title = "Spearman Correlation Coefficients mtcars1",
                                  cex.main=0.8,
                                  mar=c(0,0,2,0))

p_iris_grob <- corrplot_to_grob(cor_iris,
                                method="color",
                                addCoef.col = "black",
                                number.cex=0.6,
                                tl.cex = 0.8,
                                addgrid.col = "grey",
                                cl.cex=0.6,
                                title = "Spearman Correlation Coefficients iris",
                                cex.main=0.8,
                                mar=c(0,0,2,0))

# 合并图形
p_list <- list(p_mtcars_grob, p_iris_grob)
p_combined <- plot_grid(plotlist = p_list, ncol = 4, labels = "AUTO")
p_combined

可以通过以下几种方式导出最终的合并图：

1. 使用cowplot的save_plot() #

# 导出为PNG，可自定义尺寸和分辨率
save_plot("combined_corrplots.png", p_combined, base_width = 16, base_height = 12, dpi = 300)
# 导出为PDF
save_plot("combined_corrplots.pdf", p_combined, base_width = 16, base_height = 12)

2. 使用基础R绘图设备 #

# 导出高分辨率PNG
png("combined_corrplots.png", width = 16, height = 12, units = "in", res = 300)
print(p_combined)
dev.off()

# 导出PDF
pdf("combined_corrplots.pdf", width = 16, height = 12)
print(p_combined)
dev.off()

3. 使用patchwork的ggsave()（若用patchwork合并） #

ggsave("combined_corrplots.png", p_combined_patch, width = 16, height = 12, dpi = 300)

内容的提问来源于stack exchange，提问作者wesleysc352