介绍

ggpubr是我经常会用到的R包，它傻瓜式的画图方式对很多初次接触R绘图的人来讲是很友好的。该包有个stat_compare_means函数可以做组间假设检验分析。更多知识分享请到 [https://zouhua.top/](https://zouhua.top/**)。

安装

安装R包可参考如何安装R包。

  1. install.packages("ggpubr")
  2. devtools::devtools::install_github("kassambara/ggpubr")
  3. library(ggpubr)
  6. plotdata <- data.frame(sex = factor(rep(c("F", "M"), each=200)),
  7.                    weight = c(rnorm(200, 55), rnorm(200, 58)))

密度图density

  1. ggdensity(plotdata, 
  2.           x = "weight",
  3.           add = "mean", 
  4.           rug = TRUE,    # x轴显示分布密度
  5.           color = "sex", 
  6.           fill = "sex",
  7.           palette = c("#00AFBB", "#E7B800"))

R包：ggpubr傻瓜式绘图包 - 图1

柱状图histogram

  1. gghistogram(plotdata, 
  2.             x = "weight",
  3.             bins = 30,
  4.             add = "mean", 
  5.             rug = TRUE,
  6.             color = "sex", 
  7.             fill = "sex",
  8.             palette = c("#00AFBB", "#E7B800"))

R包：ggpubr傻瓜式绘图包 - 图2

箱线图boxplot

  1. df <- ToothGrowth
  2. head(df)
  3. my_comparisons <- list( c("0.5", "1"), c("1", "2"), c("0.5", "2") )
  4. ggboxplot(df, 
  5.           x = "dose", 
  6.           y = "len",
  7.           color = "dose", 
  8.           palette =c("#00AFBB", "#E7B800", "#FC4E07"),
  9.           add = "jitter", 
  10.           shape = "dose")+
  11.   stat_compare_means(comparisons = my_comparisons)+ # Add pairwise comparisons p-value
  12.   stat_compare_means(label.y = 50)

R包：ggpubr傻瓜式绘图包 - 图3

小提琴图violin

  1. ggviolin(df, 
  2.          x = "dose", 
  3.          y = "len", 
  4.          fill = "dose",
  5.          palette = c("#00AFBB", "#E7B800", "#FC4E07"),
  6.          add = "boxplot", 
  7.          add.params = list(fill = "white"))+
  8.   stat_compare_means(comparisons = my_comparisons, label = "p.signif")+ # Add significance levels
  9.   stat_compare_means(label.y = 50)

R包：ggpubr傻瓜式绘图包 - 图4

点图dotplot

  1. ggdotplot(ToothGrowth, 
  2.           x = "dose", 
  3.           y = "len",
  4.           color = "dose", 
  5.           palette = "jco", 
  6.           binwidth = 1)

R包：ggpubr傻瓜式绘图包 - 图5

有序条形图 ordered bar plots

  1. data("mtcars")
  2. dfm <- mtcars
  3. dfm$cyl <- as.factor(dfm$cyl)
  4. dfm$name <- rownames(dfm)
  5. head(dfm[, c("name", "wt", "mpg", "cyl")])
  7. ggbarplot(dfm, 
  8.           x = "name", y = "mpg",
  9.           fill = "cyl",               # change fill color by cyl
  10.           color = "white",            # Set bar border colors to white
  11.           palette = "jco",            # jco journal color palett. see ?ggpar
  12.           sort.val = "asc",           # Sort the value in dscending order
  13.           sort.by.groups = TRUE,      # Sort inside each group
  14.           x.text.angle = 90)          # Rotate vertically x axis texts

R包：ggpubr傻瓜式绘图包 - 图6

偏差图Deviation graphs

  1. dfm$mpg_z <- (dfm$mpg -mean(dfm$mpg))/sd(dfm$mpg)
  2. dfm$mpg_grp <- factor(ifelse(dfm$mpg_z < 0, "low", "high"), 
  3.                       levels = c("low", "high"))
  4. # Inspect the data
  5. head(dfm[, c("name", "wt", "mpg", "mpg_z", "mpg_grp", "cyl")])
  7. ggbarplot(dfm, x = "name", y = "mpg_z",
  8.           fill = "mpg_grp",           # change fill color by mpg_level
  9.           color = "white",            # Set bar border colors to white
  10.           palette = "jco",            # jco journal color palett. see ?ggpar
  11.           sort.val = "asc",           # Sort the value in ascending order
  12.           sort.by.groups = FALSE,     # Don't sort inside each group
  13.           x.text.angle = 90,          # Rotate vertically x axis texts
  14.           ylab = "MPG z-score",
  15.           rotate = FALSE,
  16.           xlab = FALSE,
  17.           legend.title = "MPG Group")

R包：ggpubr傻瓜式绘图包 - 图7

棒棒糖图 lollipop chart

  1. ggdotchart(dfm, x = "name", y = "mpg",
  2.            color = "cyl",                                # Color by groups
  3.            palette = c("#00AFBB", "#E7B800", "#FC4E07"), # Custom color palette
  4.            sorting = "descending",                       # Sort value in descending order
  5.            add = "segments",                             # Add segments from y = 0 to dots
  6.            rotate = TRUE,                                # Rotate vertically
  7.            group = "cyl",                                # Order by groups
  8.            dot.size = 6,                                 # Large dot size
  9.            label = round(dfm$mpg),                       # Add mpg values as dot labels
  10.            font.label = list(color = "white", size = 9, 
  11.                              vjust = 0.5),               # Adjust label parameters
  12.            ggtheme = theme_pubr())                       # ggplot2 theme

R包：ggpubr傻瓜式绘图包 - 图8

偏差图Deviation graph

  1. ggdotchart(dfm, x = "name", y = "mpg_z",
  2.            color = "cyl",                                # Color by groups
  3.            palette = c("#00AFBB", "#E7B800", "#FC4E07"), # Custom color palette
  4.            sorting = "descending",                       # Sort value in descending order
  5.            add = "segments",                             # Add segments from y = 0 to dots
  6.            add.params = list(color = "lightgray", size = 2), # Change segment color and size
  7.            group = "cyl",                                # Order by groups
  8.            dot.size = 6,                                 # Large dot size
  9.            label = round(dfm$mpg_z,1),                   # Add mpg values as dot labels
  10.            font.label = list(color = "white", size = 9, 
  11.                              vjust = 0.5),               # Adjust label parameters
  12.            ggtheme = theme_pubr())+                      # ggplot2 theme
  13.   geom_hline(yintercept = 0, linetype = 2, color = "lightgray")

R包：ggpubr傻瓜式绘图包 - 图9

散点图scatterplot

  1. df <- datasets::iris
  2. head(df)
  3. ggscatter(df, 
  4.           x = 'Sepal.Width', 
  5.           y = 'Sepal.Length', 
  6.           palette = 'jco', 
  7.           shape = 'Species', 
  8.           add = 'reg.line',
  9.           color = 'Species', 
  10.           conf.int = TRUE)

R包：ggpubr傻瓜式绘图包 - 图10

  • 添加回归线的系数
  1. ggscatter(df, 
  2.           x = 'Sepal.Width', 
  3.           y = 'Sepal.Length', 
  4.           palette = 'jco', 
  5.           shape = 'Species', 
  6.           add = 'reg.line',
  7.           color = 'Species', 
  8.           conf.int = TRUE)+
  9.   stat_cor(aes(color=Species),method = "pearson", label.x = 3)

R包：ggpubr傻瓜式绘图包 - 图11

  • 添加聚类椭圆 concentration ellipses
  1. data("mtcars")
  2. dfm <- mtcars
  3. dfm$cyl <- as.factor(dfm$cyl)
  4. dfm$name <- rownames(dfm)
  6. p1 <- ggscatter(dfm, 
  7.           x = "wt", 
  8.           y = "mpg",
  9.           color = "cyl", 
  10.           palette = "jco",
  11.           shape = "cyl",
  12.           ellipse = TRUE)
  13. p2 <- ggscatter(dfm, 
  14.                 x = "wt", 
  15.                 y = "mpg",
  16.                 color = "cyl", 
  17.                 palette = "jco",
  18.                 shape = "cyl",
  19.                 ellipse = TRUE,
  20.                 ellipse.type = "convex")
  21. cowplot::plot_grid(p1, p2, align = "hv", nrow = 1)

R包：ggpubr傻瓜式绘图包 - 图12

  • 添加mean和stars
  1. ggscatter(dfm, x = "wt", y = "mpg",
  2.           color = "cyl", palette = "jco",
  3.           shape = "cyl",
  4.           ellipse = TRUE, 
  5.           mean.point = TRUE,
  6.           star.plot = TRUE)

R包：ggpubr傻瓜式绘图包 - 图13

  • 显示点标签
  1. dfm$name <- rownames(dfm)
  2. p3 <- ggscatter(dfm, 
  3.           x = "wt", 
  4.           y = "mpg",
  5.           color = "cyl", 
  6.           palette = "jco",
  7.           label = "name",
  8.           repel = TRUE)
  9. p4 <- ggscatter(dfm, 
  10.                 x = "wt", 
  11.                 y = "mpg",
  12.                 color = "cyl", 
  13.                 palette = "jco",
  14.                 label = "name",
  15.                 repel = TRUE,
  16.                 label.select = c("Toyota Corolla", "Merc 280", "Duster 360"))
  17. cowplot::plot_grid(p3, p4, align = "hv", nrow = 1)

R包：ggpubr傻瓜式绘图包 - 图14

气泡图bubble plot

  1. ggscatter(dfm, 
  2.           x = "wt", 
  3.           y = "mpg",
  4.           color = "cyl",
  5.           palette = "jco",
  6.           size = "qsec", 
  7.           alpha = 0.5)+
  8.   scale_size(range = c(0.5, 15))    # Adjust the range of points size

R包：ggpubr傻瓜式绘图包 - 图15

连线图 lineplot

  1. p1 <- ggbarplot(ToothGrowth, 
  2.           x = "dose", 
  3.           y = "len", 
  4.           add = "mean_se",
  5.           color = "supp", 
  6.           palette = "jco", 
  7.           position = position_dodge(0.8))+
  8.   stat_compare_means(aes(group = supp), label = "p.signif", label.y = 29)
  9. p2 <- ggline(ToothGrowth, 
  10.        x = "dose", 
  11.        y = "len", 
  12.        add = "mean_se",
  13.        color = "supp", 
  14.        palette = "jco")+
  15.   stat_compare_means(aes(group = supp), label = "p.signif", 
  16.                      label.y = c(16, 25, 29))
  17. cowplot::plot_grid(p1, p2, ncol = 2, align = "hv")

R包：ggpubr傻瓜式绘图包 - 图16

添加边沿图 marginal plots

  1. library(ggExtra)
  2. p <- ggscatter(iris, 
  3.                x = "Sepal.Length", 
  4.                y = "Sepal.Width",
  5.                color = "Species", 
  6.                palette = "jco",
  7.                size = 3, 
  8.                alpha = 0.6)
  9. ggMarginal(p, type = "boxplot")

R包：ggpubr傻瓜式绘图包 - 图17

  • 第二种添加方式: 分别画出三个图，然后进行组合
  1. sp <- ggscatter(iris, 
  2.                 x = "Sepal.Length", 
  3.                 y = "Sepal.Width",
  4.                 color = "Species", 
  5.                 palette = "jco",
  6.                 size = 3, 
  7.                 alpha = 0.6, 
  8.                 ggtheme = theme_bw())             
  10. xplot <- ggboxplot(iris, 
  11.                    x = "Species", 
  12.                    y = "Sepal.Length", 
  13.                    color = "Species", 
  14.                    fill = "Species", 
  15.                    palette = "jco",
  16.                    alpha = 0.5, 
  17.                    ggtheme = theme_bw())+ rotate()
  19. yplot <- ggboxplot(iris, 
  20.                    x = "Species", 
  21.                    y = "Sepal.Width",
  22.                    color = "Species", 
  23.                    fill = "Species", 
  24.                    palette = "jco",
  25.                    alpha = 0.5, 
  26.                    ggtheme = theme_bw())
  29. sp <- sp + rremove("legend")
  30. yplot <- yplot + clean_theme() + rremove("legend")
  31. xplot <- xplot + clean_theme() + rremove("legend")
  32. cowplot::plot_grid(xplot, NULL, sp, yplot, ncol = 2, align = "hv", 
  33.           rel_widths = c(2, 1), rel_heights = c(1, 2))

R包：ggpubr傻瓜式绘图包 - 图18

  • 上图主图和边沿图之间的space太大，第三种方法能克服这个缺点
  1. library(cowplot) 
  2. # Main plot
  3. pmain <- ggplot(iris, aes(x = Sepal.Length, y = Sepal.Width, color = Species))+
  4.   geom_point()+
  5.   ggpubr::color_palette("jco")
  7. # Marginal densities along x axis
  8. xdens <- axis_canvas(pmain, axis = "x")+
  9.   geom_density(data = iris, aes(x = Sepal.Length, fill = Species),
  10.                alpha = 0.7, size = 0.2)+
  11.   ggpubr::fill_palette("jco")
  13. # Marginal densities along y axis
  14. # Need to set coord_flip = TRUE, if you plan to use coord_flip()
  15. ydens <- axis_canvas(pmain, axis = "y", coord_flip = TRUE)+
  16.   geom_boxplot(data = iris, aes(x = Sepal.Width, fill = Species),
  17.                alpha = 0.7, size = 0.2)+
  18.   coord_flip()+
  19.   ggpubr::fill_palette("jco")
  21. p1 <- insert_xaxis_grob(pmain, xdens, grid::unit(.2, "null"), position = "top")
  22. p2 <- insert_yaxis_grob(p1, ydens, grid::unit(.2, "null"), position = "right")
  23. ggdraw(p2)

R包：ggpubr傻瓜式绘图包 - 图19

  • 第四种方法，通过grob设置
  1. # Scatter plot colored by groups ("Species")
  2. #::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
  3. sp <- ggscatter(iris, x = "Sepal.Length", y = "Sepal.Width",
  4.                 color = "Species", palette = "jco",
  5.                 size = 3, alpha = 0.6)
  6. # Create box plots of x/y variables
  7. #::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
  8. # Box plot of the x variable
  9. xbp <- ggboxplot(iris$Sepal.Length, width = 0.3, fill = "lightgray") +
  10.   rotate() +
  11.   theme_transparent()
  12. # Box plot of the y variable
  13. ybp <- ggboxplot(iris$Sepal.Width, width = 0.3, fill = "lightgray") +
  14.   theme_transparent()
  15. # Create the external graphical objects
  16. # called a "grop" in Grid terminology
  17. xbp_grob <- ggplotGrob(xbp)
  18. ybp_grob <- ggplotGrob(ybp)
  19. # Place box plots inside the scatter plot
  20. #::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
  21. xmin <- min(iris$Sepal.Length); xmax <- max(iris$Sepal.Length)
  22. ymin <- min(iris$Sepal.Width); ymax <- max(iris$Sepal.Width)
  23. yoffset <- (1/15)*ymax; xoffset <- (1/15)*xmax
  24. # Insert xbp_grob inside the scatter plot
  25. sp + annotation_custom(grob = xbp_grob, xmin = xmin, xmax = xmax, 
  26.                        ymin = ymin-yoffset, ymax = ymin+yoffset) +
  27.   # Insert ybp_grob inside the scatter plot
  28.   annotation_custom(grob = ybp_grob,
  29.                        xmin = xmin-xoffset, xmax = xmin+xoffset, 
  30.                        ymin = ymin, ymax = ymax)

R包：ggpubr傻瓜式绘图包 - 图20

二维密度图 2d density

  1. sp <- ggscatter(iris, x = "Sepal.Length", y = "Sepal.Width",
  2.                 color = "lightgray")
  3. p1 <- sp + geom_density_2d()
  4. # Gradient color
  5. p2 <- sp + stat_density_2d(aes(fill = ..level..), geom = "polygon")
  6. # Change gradient color: custom
  7. p3 <- sp + stat_density_2d(aes(fill = ..level..), geom = "polygon")+
  8.   gradient_fill(c("white", "steelblue"))
  9. # Change the gradient color: RColorBrewer palette
  10. p4 <- sp + stat_density_2d(aes(fill = ..level..), geom = "polygon") +
  11.   gradient_fill("YlOrRd")
  13. cowplot::plot_grid(p1, p2, p3, p4, ncol = 2, align = "hv")

R包：ggpubr傻瓜式绘图包 - 图21

混合图

混合表、字体和图

  1. # Density plot of "Sepal.Length"
  2. #::::::::::::::::::::::::::::::::::::::
  3. density.p <- ggdensity(iris, x = "Sepal.Length", 
  4.                        fill = "Species", palette = "jco")
  5. # Draw the summary table of Sepal.Length
  6. #::::::::::::::::::::::::::::::::::::::
  7. # Compute descriptive statistics by groups
  8. stable <- desc_statby(iris, measure.var = "Sepal.Length",
  9.                       grps = "Species")
  10. stable <- stable[, c("Species", "length", "mean", "sd")]
  11. # Summary table plot, medium orange theme
  12. stable.p <- ggtexttable(stable, rows = NULL, 
  13.                         theme = ttheme("mOrange"))
  14. # Draw text
  15. #::::::::::::::::::::::::::::::::::::::
  16. text <- paste("iris data set gives the measurements in cm",
  17.               "of the variables sepal length and width",
  18.               "and petal length and width, respectively,",
  19.               "for 50 flowers from each of 3 species of iris.",
  20.              "The species are Iris setosa, versicolor, and virginica.", sep = " ")
  21. text.p <- ggparagraph(text = text, face = "italic", size = 11, color = "black")
  22. # Arrange the plots on the same page
  23. ggarrange(density.p, stable.p, text.p, 
  24.           ncol = 1, nrow = 3,
  25.           heights = c(1, 0.5, 0.3))

R包：ggpubr傻瓜式绘图包 - 图22

  • 注释table在图上
  1. density.p <- ggdensity(iris, x = "Sepal.Length", 
  2.                        fill = "Species", palette = "jco")
  4. stable <- desc_statby(iris, measure.var = "Sepal.Length",
  5.                       grps = "Species")
  6. stable <- stable[, c("Species", "length", "mean", "sd")]
  7. stable.p <- ggtexttable(stable, rows = NULL, 
  8.                         theme = ttheme("mOrange"))
  9. density.p + annotation_custom(ggplotGrob(stable.p),
  10.                               xmin = 5.5, ymin = 0.7,
  11.                               xmax = 8)

R包：ggpubr傻瓜式绘图包 - 图23

systemic information

  1. sessionInfo()
  1. R version 3.6.1 (2019-07-05)
  2. Platform: x86_64-w64-mingw32/x64 (64-bit)
  3. Running under: Windows 10 x64 (build 19042)
  5. Matrix products: default
  7. locale:
  8. [1] LC_COLLATE=Chinese (Simplified)_China.936  LC_CTYPE=Chinese (Simplified)_China.936   
  9. [3] LC_MONETARY=Chinese (Simplified)_China.936 LC_NUMERIC=C                              
  10. [5] LC_TIME=Chinese (Simplified)_China.936    
  12. attached base packages:
  13. [1] stats     graphics  grDevices utils     datasets  methods   base     
  15. other attached packages:
  16. [1] ggpubr_0.4.0  ggplot2_3.3.2
  18. loaded via a namespace (and not attached):
  19.  [1] zip_2.0.4         Rcpp_1.0.3        cellranger_1.1.0  pillar_1.4.6      compiler_3.6.1    forcats_0.5.0    
  20.  [7] tools_3.6.1       digest_0.6.27     lifecycle_0.2.0   tibble_3.0.4      gtable_0.3.0      pkgconfig_2.0.3  
  21. [13] rlang_0.4.8       openxlsx_4.2.3    ggsci_2.9         rstudioapi_0.10   curl_4.3          haven_2.3.1      
  22. [19] rio_0.5.16        withr_2.1.2       dplyr_1.0.2       generics_0.0.2    vctrs_0.3.4       hms_0.5.3        
  23. [25] grid_3.6.1        tidyselect_1.1.0  glue_1.4.2        data.table_1.13.2 R6_2.4.1          rstatix_0.6.0    
  24. [31] readxl_1.3.1      foreign_0.8-73    carData_3.0-4     farver_2.0.3      tidyr_1.0.0       purrr_0.3.3      
  25. [37] car_3.0-10        magrittr_1.5      scales_1.1.0      backports_1.1.10  ellipsis_0.3.1    abind_1.4-5      
  26. [43] colorspace_1.4-1  ggsignif_0.6.0    labeling_0.4.2    stringi_1.4.3     munsell_0.5.0     broom_0.7.2      
  27. [49] crayon_1.3.4

Reference

  1. ggpubr: Publication Ready Plots

参考文章如引起任何侵权问题，可以与我联系，谢谢。