I am using facet_wrap and was also able to plot the secondary y-axis. However the labels are not getting plotted near the axis, rather they are plotted very far. I realise it all will get resolved if I understand how to manipulate the coordinate system of the gtable (t,b,l,r) of the grobs. Could someone explain how and what they actually depict - t:r = c(4,8,4,4) means what.
我使用的是facet_wrap，也可以绘制次y轴。然而，这些标签并不是在轴线附近绘制的，而是绘制得非常远。我意识到，如果我理解了如何操纵grobs的gtable (t,b,l,r)的坐标系统，一切都会得到解决。有人能解释一下它们的真实描述吗? t:r = c(4,8,4,4)。
There are many links for secondary yaxis with ggplot, however when nrow/ncol is more than 1, they fails. So please teach me the basics of grid geometry and grobs location management.
当nrow/ncol大于1时，有很多次yaxis与ggplot的链接，它们失败了。所以请教我网格几何的基础知识和grobs位置管理。
Edit : the Code
编辑:代码
this is the final code written by me :
library(ggplot2)
library(gtable)
library(grid)
library(data.table)
library(scales)
# Data
diamonds$cut <- sample(letters[1:13], nrow(diamonds), replace = TRUE)
dt.diamonds <- as.data.table(diamonds)
d1 <- dt.diamonds[,list(revenue = sum(price),
stones = length(price)),
by=c("clarity", "cut")]
setkey(d1, clarity, cut)
# The facet_wrap plots
p1 <- ggplot(d1, aes(x = clarity, y = revenue, fill = cut)) +
geom_bar(stat = "identity") +
labs(x = "clarity", y = "revenue") +
facet_wrap( ~ cut) +
scale_y_continuous(labels = dollar, expand = c(0, 0)) +
theme(axis.text.x = element_text(angle = 90, hjust = 1),
axis.text.y = element_text(colour = "#4B92DB"),
legend.position = "bottom")
p2 <- ggplot(d1, aes(x = clarity, y = stones, colour = "red")) +
geom_point(size = 4) +
labs(x = "", y = "number of stones") + expand_limits(y = 0) +
scale_y_continuous(labels = comma, expand = c(0, 0)) +
scale_colour_manual(name = '', values = c("red", "green"),
labels = c("Number of Stones"))+
facet_wrap( ~ cut) +
theme(axis.text.y = element_text(colour = "red")) +
theme(panel.background = element_rect(fill = NA),
panel.grid.major = element_blank(),
panel.grid.minor = element_blank(),
panel.border = element_rect(fill = NA, colour = "grey50"),
legend.position = "bottom")
# Get the ggplot grobs
xx <- ggplot_build(p1)
g1 <- ggplot_gtable(xx)
yy <- ggplot_build(p2)
g2 <- ggplot_gtable(yy)
nrow = length(unique(xx$panel$layout$ROW))
ncol = length(unique(xx$panel$layout$COL))
npanel = length(xx$panel$layout$PANEL)
pp <- c(subset(g1$layout, grepl("panel", g1$layout$name), se = t:r))
g <- gtable_add_grob(g1, g2$grobs[grepl("panel", g1$layout$name)],
pp$t, pp$l, pp$b, pp$l)
hinvert_title_grob <- function(grob){
widths <- grob$widths
grob$widths[1] <- widths[3]
grob$widths[3] <- widths[1]
grob$vp[[1]]$layout$widths[1] <- widths[3]
grob$vp[[1]]$layout$widths[3] <- widths[1]
grob$children[[1]]$hjust <- 1 - grob$children[[1]]$hjust
grob$children[[1]]$vjust <- 1 - grob$children[[1]]$vjust
grob$children[[1]]$x <- unit(1, "npc") - grob$children[[1]]$x
grob
}
j = 1
k = 0
for(i in 1:npanel){
if ((i %% ncol == 0) || (i == npanel)){
k = k + 1
index <- which(g2$layout$name == "axis_l-1") # Which grob
yaxis <- g2$grobs[[index]] # Extract the grob
ticks <- yaxis$children[[2]]
ticks$widths <- rev(ticks$widths)
ticks$grobs <- rev(ticks$grobs)
ticks$grobs[[1]]$x <- ticks$grobs[[1]]$x - unit(1, "npc")
ticks$grobs[[2]] <- hinvert_title_grob(ticks$grobs[[2]])
yaxis$children[[2]] <- ticks
if (k == 1)#to ensure just once d secondary axisis printed
g <- gtable_add_cols(g,g2$widths[g2$layout[index,]$l],
max(pp$r[j:i]))
g <- gtable_add_grob(g,yaxis,max(pp$t[j:i]),max(pp$r[j:i])+1,
max(pp$b[j:i])
, max(pp$r[j:i]) + 1, clip = "off", name = "2ndaxis")
j = i + 1
}
}
# inserts the label for 2nd y-axis
loc_1st_yaxis_label <- c(subset(g$layout, grepl("ylab", g$layout$name), se
= t:r))
loc_2nd_yaxis_max_r <- c(subset(g$layout, grepl("2ndaxis", g$layout$name),
se = t:r))
zz <- max(loc_2nd_yaxis_max_r$r)+1
loc_1st_yaxis_label$l <- zz
loc_1st_yaxis_label$r <- zz
index <- which(g2$layout$name == "ylab")
ylab <- g2$grobs[[index]] # Extract that grob
ylab <- hinvert_title_grob(ylab)
ylab$children[[1]]$rot <- ylab$children[[1]]$rot + 180
g <- gtable_add_grob(g, ylab, loc_1st_yaxis_label$t, loc_1st_yaxis_label$l
, loc_1st_yaxis_label$b, loc_1st_yaxis_label$r
, clip = "off", name = "2ndylab")
grid.draw(g)
@Sandy here is the code and its output
这里是代码和它的输出。
only trouble was that in the last row the secondary y-axis labels are inside the panels.I tried to solve this but not able to
唯一的问题是，在最后一行中，次级y轴标签在面板中。我试着解决这个问题，但没能解决。
1 个解决方案
#1
10
There were problems with your gtable_add_cols() and gtable_add_grob() commands. I added comments below.
您的gtable_add_cols()和gtable_add_grob()命令存在问题。我添加了下面的评论。
Updated to ggplot2 v2.2.0
更新为ggplot2 v2.2.0
library(ggplot2)
library(gtable)
library(grid)
library(data.table)
library(scales)
diamonds$cut <- sample(letters[1:4], nrow(diamonds), replace = TRUE)
dt.diamonds <- as.data.table(diamonds)
d1 <- dt.diamonds[,list(revenue = sum(price),
stones = length(price)),
by=c("clarity", "cut")]
setkey(d1, clarity, cut)
# The facet_wrap plots
p1 <- ggplot(d1, aes(x = clarity, y = revenue, fill = cut)) +
geom_bar(stat = "identity") +
labs(x = "clarity", y = "revenue") +
facet_wrap( ~ cut, nrow = 2) +
scale_y_continuous(labels = dollar, expand = c(0, 0)) +
theme(axis.text.x = element_text(angle = 90, hjust = 1),
axis.text.y = element_text(colour = "#4B92DB"),
legend.position = "bottom")
p2 <- ggplot(d1, aes(x = clarity, y = stones, colour = "red")) +
geom_point(size = 4) +
labs(x = "", y = "number of stones") + expand_limits(y = 0) +
scale_y_continuous(labels = comma, expand = c(0, 0)) +
scale_colour_manual(name = '', values = c("red", "green"),
labels =c("Number of Stones")) +
facet_wrap( ~ cut, nrow = 2) +
theme(axis.text.y = element_text(colour = "red")) +
theme(panel.background = element_rect(fill = NA),
panel.grid.major = element_blank(),
panel.grid.minor = element_blank(),
panel.border = element_rect(fill = NA, colour = "grey50"),
legend.position = "bottom")
# Get the ggplot grobs
g1 <- ggplotGrob(p1)
g2 <- ggplotGrob(p2)
# Grab the panels from g2 and overlay them onto the panels of g1
pp <- c(subset(g1$layout, grepl("panel", g1$layout$name), select = t:r))
g <- gtable_add_grob(g1, g2$grobs[grepl("panel", g1$layout$name)],
pp$t, pp$l, pp$b, pp$l)
# Function to invert labels
hinvert_title_grob <- function(grob){
widths <- grob$widths
grob$widths[1] <- widths[3]
grob$widths[3] <- widths[1]
grob$vp[[1]]$layout$widths[1] <- widths[3]
grob$vp[[1]]$layout$widths[3] <- widths[1]
grob$children[[1]]$hjust <- 1 - grob$children[[1]]$hjust
grob$children[[1]]$vjust <- 1 - grob$children[[1]]$vjust
grob$children[[1]]$x <- unit(1, "npc") - grob$children[[1]]$x
grob
}
# Get the y label from g2, and invert it
index <- which(g2$layout$name == "ylab-l")
ylab <- g2$grobs[[index]] # Extract that grob
ylab <- hinvert_title_grob(ylab)
# Put the y label into g, to the right of the right-most panel
# Note: Only one column and one y label
g <- gtable_add_cols(g, g2$widths[g2$layout[index, ]$l], pos = max(pp$r))
g <-gtable_add_grob(g,ylab, t = min(pp$t), l = max(pp$r)+1,
b = max(pp$b), r = max(pp$r)+1,
clip = "off", name = "ylab-r")
# Get the y axis from g2, reverse the tick marks and the tick mark labels,
# and invert the tick mark labels
index <- which(g2$layout$name == "axis-l-1-1") # Which grob
yaxis <- g2$grobs[[index]] # Extract the grob
ticks <- yaxis$children[[2]]
ticks$widths <- rev(ticks$widths)
ticks$grobs <- rev(ticks$grobs)
plot_theme <- function(p) {
plyr::defaults(p$theme, theme_get())
}
tml <- plot_theme(p1)$axis.ticks.length # Tick mark length
ticks$grobs[[1]]$x <- ticks$grobs[[1]]$x - unit(1, "npc") + tml
ticks$grobs[[2]] <- hinvert_title_grob(ticks$grobs[[2]])
yaxis$children[[2]] <- ticks
# Put the y axis into g, to the right of the right-most panel
# Note: Only one column, but two y axes - one for each row of the facet_wrap plot
g <- gtable_add_cols(g, g2$widths[g2$layout[index, ]$l], pos = max(pp$r))
nrows = length(unique(pp$t)) # Number of rows
g <- gtable_add_grob(g, rep(list(yaxis), nrows),
t = unique(pp$t), l = max(pp$r)+1,
b = unique(pp$b), r = max(pp$r)+1,
clip = "off", name = paste0("axis-r-", 1:nrows))
# Get the legends
leg1 <- g1$grobs[[which(g1$layout$name == "guide-box")]]
leg2 <- g2$grobs[[which(g2$layout$name == "guide-box")]]
# Combine the legends
g$grobs[[which(g$layout$name == "guide-box")]] <-
gtable:::cbind_gtable(leg1, leg2, "first")
grid.newpage()
grid.draw(g)
SO is not a tutorial site, and this might incur the wrath of other SO users, but there is too much for a comment.
因此，这不是一个教程网站，这可能招致其他用户的愤怒，但是有太多的评论。
Draw a graph with one plot panel only (i.e., no facetting),
只画一个图，只画一个图。,没有方面),
library(ggplot2)
p <- ggplot(mtcars, aes(x = mpg, y = disp)) + geom_point()
Get the ggplot grob.
grob ggplot。
g <- ggplotGrob(p)
Explore the plot grob:
1) gtable_show_layout() give a diagram of the plot's gtable layout. The big space in the middle is the location of the plot panel. Columns to the left of and below the panel contain the y and x axes. And there is a margin surrounding the whole plot. The indices give the location of each cell in the array. Note, for instance, that the panel is located in the third row of the fourth column.
查看plot grob: 1) gtable_show_layout()给出plot的gtable布局图。中间的大空间是plot面板的位置。面板左侧和下方的列包含y轴和x轴。整个剧情围绕着一个边缘。索引给出数组中每个单元格的位置。例如，注意面板位于第四列的第三行。
gtable_show_layout(g)
2) The layout dataframe. g$layout returns a dataframe which contains the names of the grobs contained in the plot along with their locations within the gtable: t, l, b, and r (standing for top, left, right, and bottom). Note, for instance, that the panel is located at t=3, l=4, b=3, r=4. That is the same panel location that was obtained above from the diagram.
2)dataframe布局。g$layout返回一个dataframe，其中包含该绘图中包含的grobs的名称及其在gtable中的位置:t、l、b和r(分别位于顶部、左、右和底部)。注意，例如，面板位于t=3, l=4, b=3, r=4。这是上图所示的面板位置。
g$layout
3) The diagram of the layout tries to give the heights and widths of the rows and columns, but they tend to overlap. Instead, use g$widths and g$heights. The 1null width and height is the width and height of the plot panel. Note that 1null is the 3rd height and the 4th width - 3 and 4 again.
3)布局图试图给出行和列的高度和宽度，但它们趋向于重叠。相反，使用g$widths和g$heights。1null宽度和高度是plot面板的宽度和高度。注意，1null是第3个高度，第4个宽度是3和4。
Now draw a facet_wrap and a facet_grid plot.
现在绘制facet_wrap和facet_grid图。
p1 <- ggplot(mtcars, aes(x = mpg, y = disp)) + geom_point() +
facet_wrap(~ carb, nrow = 1)
p2 <- ggplot(mtcars, aes(x = mpg, y = disp)) + geom_point() +
facet_grid(. ~ carb)
g1 <- ggplotGrob(p1)
g2 <- ggplotGrob(p2)
The two plots look the same, but their gtables differ. Also, the names of the component grobs differ.
这两个图看起来是一样的，但是它们的gtables是不同的。同时，组件的名称也不同。
Often it is convenient to get a subset of the layout dataframe containing the indices (i.e., t, l, b, and r) of grobs of a common type; say all the panels.
通常，获取包含索引的布局dataframe的子集是很方便的。一种常见类型的grobs的t、l、b、r);说所有的面板。
pp1 <- subset(g1$layout, grepl("panel", g1$layout$name), select = t:r)
pp2 <- subset(g2$layout, grepl("panel", g2$layout$name), select = t:r)
Note for instance that all the panels are in row 4 (pp1$t, pp2$t).pp1$r refers to the columns that contain the plot panels;pp1$r + 1 refers to the columns to the right of the panels;max(pp1$r) refers to the right most column that contains a panel;max(pp1$r) + 1 refers to the column to the right of the right most column that contains a panel;
and so forth.
例如，所有的面板都在第4行(pp1$t, pp2$t)。pp1$r是指包含图表面板的列;pp1$r + 1是指面板右侧的列;max(pp1$r)是指包含一个面板的最右边的列;max(pp1$r) + 1是指右边最右边的列，其中包含一个面板;等等。
Finally, draw a facet_wrap plot with more than one row.
最后，绘制一个包含不止一行的facet_wrap图。
p3 <- ggplot(mtcars, aes(x = mpg, y = disp)) + geom_point() +
facet_wrap(~ carb, nrow = 2)
g3 <- ggplotGrob(p3)
Explore the plot as before, but also subset the layout data frame to contain the indices of the panels.
像以前一样探索这个图，但也要对布局数据帧进行子集以包含面板的索引。
pp3 <- subset(g3$layout, grepl("panel", g3$layout$name), select = t:r)
As you would expect, pp3 tells you that the plot panels are located in three columns (4, 7, and 10) and two rows (4 and 8).
正如您所期望的，pp3告诉您plot面板位于三列(4、7和10)和2行(4和8)。
These indices are used when adding rows or columns to the gtable, and when adding grobs to a gtable. Check these commands with ?gtable_add_rows and gtable_add_grob.
在向gtable中添加行或列时使用这些索引，并将grobs添加到gtable中。使用gtable_add_rows和gtable_add_grob检查这些命令。
Also, learn some grid, especially how to construct grobs, and the use of units (some resources are given in the r-grid tag here on SO.
另外，学习一些网格，特别是如何构建grobs，以及单元的使用(这里的r-grid标记中给出了一些资源)。
#1
10
There were problems with your gtable_add_cols() and gtable_add_grob() commands. I added comments below.
您的gtable_add_cols()和gtable_add_grob()命令存在问题。我添加了下面的评论。
Updated to ggplot2 v2.2.0
更新为ggplot2 v2.2.0
library(ggplot2)
library(gtable)
library(grid)
library(data.table)
library(scales)
diamonds$cut <- sample(letters[1:4], nrow(diamonds), replace = TRUE)
dt.diamonds <- as.data.table(diamonds)
d1 <- dt.diamonds[,list(revenue = sum(price),
stones = length(price)),
by=c("clarity", "cut")]
setkey(d1, clarity, cut)
# The facet_wrap plots
p1 <- ggplot(d1, aes(x = clarity, y = revenue, fill = cut)) +
geom_bar(stat = "identity") +
labs(x = "clarity", y = "revenue") +
facet_wrap( ~ cut, nrow = 2) +
scale_y_continuous(labels = dollar, expand = c(0, 0)) +
theme(axis.text.x = element_text(angle = 90, hjust = 1),
axis.text.y = element_text(colour = "#4B92DB"),
legend.position = "bottom")
p2 <- ggplot(d1, aes(x = clarity, y = stones, colour = "red")) +
geom_point(size = 4) +
labs(x = "", y = "number of stones") + expand_limits(y = 0) +
scale_y_continuous(labels = comma, expand = c(0, 0)) +
scale_colour_manual(name = '', values = c("red", "green"),
labels =c("Number of Stones")) +
facet_wrap( ~ cut, nrow = 2) +
theme(axis.text.y = element_text(colour = "red")) +
theme(panel.background = element_rect(fill = NA),
panel.grid.major = element_blank(),
panel.grid.minor = element_blank(),
panel.border = element_rect(fill = NA, colour = "grey50"),
legend.position = "bottom")
# Get the ggplot grobs
g1 <- ggplotGrob(p1)
g2 <- ggplotGrob(p2)
# Grab the panels from g2 and overlay them onto the panels of g1
pp <- c(subset(g1$layout, grepl("panel", g1$layout$name), select = t:r))
g <- gtable_add_grob(g1, g2$grobs[grepl("panel", g1$layout$name)],
pp$t, pp$l, pp$b, pp$l)
# Function to invert labels
hinvert_title_grob <- function(grob){
widths <- grob$widths
grob$widths[1] <- widths[3]
grob$widths[3] <- widths[1]
grob$vp[[1]]$layout$widths[1] <- widths[3]
grob$vp[[1]]$layout$widths[3] <- widths[1]
grob$children[[1]]$hjust <- 1 - grob$children[[1]]$hjust
grob$children[[1]]$vjust <- 1 - grob$children[[1]]$vjust
grob$children[[1]]$x <- unit(1, "npc") - grob$children[[1]]$x
grob
}
# Get the y label from g2, and invert it
index <- which(g2$layout$name == "ylab-l")
ylab <- g2$grobs[[index]] # Extract that grob
ylab <- hinvert_title_grob(ylab)
# Put the y label into g, to the right of the right-most panel
# Note: Only one column and one y label
g <- gtable_add_cols(g, g2$widths[g2$layout[index, ]$l], pos = max(pp$r))
g <-gtable_add_grob(g,ylab, t = min(pp$t), l = max(pp$r)+1,
b = max(pp$b), r = max(pp$r)+1,
clip = "off", name = "ylab-r")
# Get the y axis from g2, reverse the tick marks and the tick mark labels,
# and invert the tick mark labels
index <- which(g2$layout$name == "axis-l-1-1") # Which grob
yaxis <- g2$grobs[[index]] # Extract the grob
ticks <- yaxis$children[[2]]
ticks$widths <- rev(ticks$widths)
ticks$grobs <- rev(ticks$grobs)
plot_theme <- function(p) {
plyr::defaults(p$theme, theme_get())
}
tml <- plot_theme(p1)$axis.ticks.length # Tick mark length
ticks$grobs[[1]]$x <- ticks$grobs[[1]]$x - unit(1, "npc") + tml
ticks$grobs[[2]] <- hinvert_title_grob(ticks$grobs[[2]])
yaxis$children[[2]] <- ticks
# Put the y axis into g, to the right of the right-most panel
# Note: Only one column, but two y axes - one for each row of the facet_wrap plot
g <- gtable_add_cols(g, g2$widths[g2$layout[index, ]$l], pos = max(pp$r))
nrows = length(unique(pp$t)) # Number of rows
g <- gtable_add_grob(g, rep(list(yaxis), nrows),
t = unique(pp$t), l = max(pp$r)+1,
b = unique(pp$b), r = max(pp$r)+1,
clip = "off", name = paste0("axis-r-", 1:nrows))
# Get the legends
leg1 <- g1$grobs[[which(g1$layout$name == "guide-box")]]
leg2 <- g2$grobs[[which(g2$layout$name == "guide-box")]]
# Combine the legends
g$grobs[[which(g$layout$name == "guide-box")]] <-
gtable:::cbind_gtable(leg1, leg2, "first")
grid.newpage()
grid.draw(g)
SO is not a tutorial site, and this might incur the wrath of other SO users, but there is too much for a comment.
因此，这不是一个教程网站，这可能招致其他用户的愤怒，但是有太多的评论。
Draw a graph with one plot panel only (i.e., no facetting),
只画一个图，只画一个图。,没有方面),
library(ggplot2)
p <- ggplot(mtcars, aes(x = mpg, y = disp)) + geom_point()
Get the ggplot grob.
grob ggplot。
g <- ggplotGrob(p)
Explore the plot grob:
1) gtable_show_layout() give a diagram of the plot's gtable layout. The big space in the middle is the location of the plot panel. Columns to the left of and below the panel contain the y and x axes. And there is a margin surrounding the whole plot. The indices give the location of each cell in the array. Note, for instance, that the panel is located in the third row of the fourth column.
查看plot grob: 1) gtable_show_layout()给出plot的gtable布局图。中间的大空间是plot面板的位置。面板左侧和下方的列包含y轴和x轴。整个剧情围绕着一个边缘。索引给出数组中每个单元格的位置。例如，注意面板位于第四列的第三行。
gtable_show_layout(g)
2) The layout dataframe. g$layout returns a dataframe which contains the names of the grobs contained in the plot along with their locations within the gtable: t, l, b, and r (standing for top, left, right, and bottom). Note, for instance, that the panel is located at t=3, l=4, b=3, r=4. That is the same panel location that was obtained above from the diagram.
2)dataframe布局。g$layout返回一个dataframe，其中包含该绘图中包含的grobs的名称及其在gtable中的位置:t、l、b和r(分别位于顶部、左、右和底部)。注意，例如，面板位于t=3, l=4, b=3, r=4。这是上图所示的面板位置。
g$layout
3) The diagram of the layout tries to give the heights and widths of the rows and columns, but they tend to overlap. Instead, use g$widths and g$heights. The 1null width and height is the width and height of the plot panel. Note that 1null is the 3rd height and the 4th width - 3 and 4 again.
3)布局图试图给出行和列的高度和宽度，但它们趋向于重叠。相反，使用g$widths和g$heights。1null宽度和高度是plot面板的宽度和高度。注意，1null是第3个高度，第4个宽度是3和4。
Now draw a facet_wrap and a facet_grid plot.
现在绘制facet_wrap和facet_grid图。
p1 <- ggplot(mtcars, aes(x = mpg, y = disp)) + geom_point() +
facet_wrap(~ carb, nrow = 1)
p2 <- ggplot(mtcars, aes(x = mpg, y = disp)) + geom_point() +
facet_grid(. ~ carb)
g1 <- ggplotGrob(p1)
g2 <- ggplotGrob(p2)
The two plots look the same, but their gtables differ. Also, the names of the component grobs differ.
这两个图看起来是一样的，但是它们的gtables是不同的。同时，组件的名称也不同。
Often it is convenient to get a subset of the layout dataframe containing the indices (i.e., t, l, b, and r) of grobs of a common type; say all the panels.
通常，获取包含索引的布局dataframe的子集是很方便的。一种常见类型的grobs的t、l、b、r);说所有的面板。
pp1 <- subset(g1$layout, grepl("panel", g1$layout$name), select = t:r)
pp2 <- subset(g2$layout, grepl("panel", g2$layout$name), select = t:r)
Note for instance that all the panels are in row 4 (pp1$t, pp2$t).pp1$r refers to the columns that contain the plot panels;pp1$r + 1 refers to the columns to the right of the panels;max(pp1$r) refers to the right most column that contains a panel;max(pp1$r) + 1 refers to the column to the right of the right most column that contains a panel;
and so forth.
例如，所有的面板都在第4行(pp1$t, pp2$t)。pp1$r是指包含图表面板的列;pp1$r + 1是指面板右侧的列;max(pp1$r)是指包含一个面板的最右边的列;max(pp1$r) + 1是指右边最右边的列，其中包含一个面板;等等。
Finally, draw a facet_wrap plot with more than one row.
最后，绘制一个包含不止一行的facet_wrap图。
p3 <- ggplot(mtcars, aes(x = mpg, y = disp)) + geom_point() +
facet_wrap(~ carb, nrow = 2)
g3 <- ggplotGrob(p3)
Explore the plot as before, but also subset the layout data frame to contain the indices of the panels.
像以前一样探索这个图，但也要对布局数据帧进行子集以包含面板的索引。
pp3 <- subset(g3$layout, grepl("panel", g3$layout$name), select = t:r)
As you would expect, pp3 tells you that the plot panels are located in three columns (4, 7, and 10) and two rows (4 and 8).
正如您所期望的，pp3告诉您plot面板位于三列(4、7和10)和2行(4和8)。
These indices are used when adding rows or columns to the gtable, and when adding grobs to a gtable. Check these commands with ?gtable_add_rows and gtable_add_grob.
在向gtable中添加行或列时使用这些索引，并将grobs添加到gtable中。使用gtable_add_rows和gtable_add_grob检查这些命令。
Also, learn some grid, especially how to construct grobs, and the use of units (some resources are given in the r-grid tag here on SO.
另外，学习一些网格，特别是如何构建grobs，以及单元的使用(这里的r-grid标记中给出了一些资源)。