[合集]R语言绘图（ggplot2，ggpubr）从入门到精通

# * 第一章：快速探索数据 ------------------------------------------------------------

###################

## 散点图

# 最简单的函数:plot(x, y)

plot(mtcars$wt,mtcars$mpg)

# 使用qplot

library(ggplot2)

qplot(mtcars$wt,mtcars$mpg) 

qplot(wt, mpg, data = mtcars)

# 使用ggplot2

ggplot(mtcars, aes(x = wt, y = mpg)) + 

 geom_point()

###################

## 折线图、曲线图

# 同样可以用plot来画:

# plot中type参数可以指定绘图的类型:如:"7"就是指折、曲线图

  

plot(pressure$temperature, pressure$pressure, type = "l") 

points (pressure$temperature, pressure$pressure)

lines(pressure$temperature, pressure$pressure/2, col ="red") 

points(pressure$temperature, pressure$pressure/2, col = "red")

# 使用qplot

library (ggplot2)

qplot (pressure$temperature, pressure$pressure, geom = "line")

qplot(temperature, pressure,data = pressure, geom = "line") #如果变量x和y 都来自于同一个数据框,还可以这样写；

# 使用ggplot2

ggplot(pressure, aes(x = temperature, y = pressure)) + 

 geom_line() +

 geom_point()

###################

## 柱状图

#使用barplot

BOD<- BOD

mtcars <- mtcars

barplot(BOD$demand,names.arg = BOD$Time)

# 有时“条形图”指的是一个图表，其中的条形图代表列每个类型的案例数

# 这类似与直方图，但是x轴是离散的，而不是连续的，这个时候要用table函数生成每个类别的计数

barplot(mtcars$cyl)

barplot(table(mtcars$cyl))

# 使用qplot()

library(ggplot2)

qplot(mtcars$cyl) #连续变量

qplot(factor(mtcars$cyl)) #分类变量

# ggplot

ggplot(mtcars,aes(cyl))+geom_bar()

# stat = "count"(默认)：表示一个x对应落到该x的样本数

# stat = "identity": 表示一个x对应一个y

# 说白了就是，identity提取横坐标x对应的y值，count提取横坐标的频数

ggplot(BOD,aes(Time,demand)) + geom_bar(stat = "identity")

ggplot(BOD,aes(factor(Time),demand)) + geom_bar(stat = "identity")

###################

## 直方图

# x是连续变量

#基础方法：hist函数

hist(mtcars$mpg)

hist(mtcars$mpg, breaks = 10)

#qplot()

qplot(mpg,data = mtcars, binwidth = 1 ) #binwidth参数指定组距

#等价于

ggplot(data = mtcars, aes(x = mpg)) + geom_histogram( binwidth = 2)

###################

## 箱线图

#使用plot（）函数，当x为因子变量（与数值变量对应时），默认绘制箱线图

ToothGrowth <- ToothGrowth

plot(ToothGrowth$supp, ToothGrowth$len)

#使用公式语法

boxplot(len ~ supp, data = ToothGrowth)

#在x轴引入两变量的交互，可以绘制分组箱线图

boxplot(len ~ supp + dose, data = ToothGrowth)

#qplot()绘制箱线图

qplot(supp,len,data = ToothGrowth, geom = "boxplot")

#使用三个独立的向量参数

qplot(interaction(supp,dose),len, data = ToothGrowth,geom = "boxplot")

#等价与

ggplot(data = ToothGrowth, aes(x = interaction(supp,dose), y = len)) + geom_boxplot()

###################

## 绘制函数图像

#使用curve()函数绘制，传入一个关于变量x的表达式

curve(x^3 - 5*x, from = -4, to =4)

#自定义函数图像

my_fun <- function(xvar){

 1/(1 + exp(-xvar + 10))

}

curve(my_fun(x), from = 0, to =20)

#原有的基础上添加一条线

curve(1 - my_fun(x),add = T, col = "red")

#等价于

ggplot(data.frame(x = c(0,20)),aes(x =x)) + stat_function(fun = my_fun, geom ="line")

my_fun2 <- function(xvar){

 1- 1/(1 + exp(-xvar + 10))

}

ggplot(data.frame(x = c(0,20)),aes(x =x)) + stat_function(fun = my_fun, geom ="line") + stat_function(fun = my_fun2, geom ="line",color="red")

# * 第二章：柱状图深入研究 ------------------------------------------------------------

library(gcookbook)

pg_mean <- pg_mean

# 修改柱状图的填充（fill）和描边（color）：

ggplot(pg_mean,aes(group, weight)) +

 geom_bar(stat = "identity", fill = "lightblue", color = "black")

# 也可通过分组变量设置颜色：

cabbage_exp <- cabbage_exp

ggplot(cabbage_exp, aes(Date, fill = Cultivar)) +

 geom_bar(position = "dodge")

 #拓展：postition参数： 主要是指对图像的微调，最常见的应用是在分组的柱形图（bar）中，因为分组的柱形图会产生祖内堆积和不堆积两种主要的效果

 # - position常用参数值：其中stack和dodge最为常用。

 # - "identity": 不调整，组内前后重叠；

 # - "stack": 堆积， 默认；

 # - "fill": 按比例堆积；

 # - "dodge": 分散  

 # - ColorBrewer 配色，使用的是scale_colour_brewer() 、scale_fill_brewer()。 想要了解所有的调色板，可以使用RColorBrewer::display.brewer.all()查看

# 通过scale_fill_brewer()修改颜色模式：

ggplot(cabbage_exp, aes(Date, Weight, fill = Cultivar)) +

 geom_bar(position = "dodge", stat = "identity", color = "black") +

 scale_fill_brewer(palette = "Pastel1")

# 可以使用以下代码查看调色板

RColorBrewer::display.brewer.all()

###################

## 分组柱状图

library(gcookbook)

cabbage_exp <- cabbage_exp

ggplot(cabbage_exp, aes(Date, Weight, fill = Cultivar)) +

 geom_bar(position = "dodge", stat = "identity", color = "black")

ggplot(cabbage_exp, aes(Date, Weight, fill = Cultivar)) + 

 geom_bar(stat = "identity") # 默认的position = "stack";

# 此外，除了fill可以指定分组变量，color, linestyle也可以指定；

ggplot(cabbage_exp, aes(Date, Weight, color = Cultivar)) + 

 geom_bar(position = "dodge", stat = "identity")

# 请注意，如果类别变量的组合有任何缺失，则该栏将缺失，相邻的栏将扩展

# 以填充该空间。

ce = cabbage_exp[1:5,]

ggplot(ce, aes(Date, Weight, fill = Cultivar)) +

 geom_bar(position = "dodge", stat = "identity", color = "black") +

 scale_fill_brewer(palette = "Pastel1")

# 实际情况下确实存在有一种类别没有对应的y值，此时可以使用NA或者0代替

ce_NA <- cabbage_exp

ce_NA$Weight[6] <- 0

ggplot(ce_NA, aes(Date, Weight, fill = Cultivar)) +

 geom_bar(position = "dodge", stat = "identity", color = "black") +

 scale_fill_brewer(palette = "Pastel1")

###################

## 修改颜色的技巧

library(gcookbook)

upc <- subset(uspopchange, rank(Change) > 40)

upc

ggplot(upc, aes(x = Abb, y = Change, fill = Region)) + 

 geom_bar(stat = "identity")

# 此示例还使用reorder()函数，将条形按其高度进行排序：

ggplot(upc, aes(x = reorder(Abb, change), y = Change, fill = Region)) +  

 geom_bar(stat = "identity", color = "black") +

 scale_fill_manual(values = c("#669933", "#FFCC66")) + 

 xlab("State")

  #拓展：reorder()函数： 

  # reorder(x, X, FUN = mean, ..., order = is.ordered(x), decreasing = FALSE)

  # x: x是要排序的数据，排序后的结果作用于x上

  # X: 要根据X进行排序

  #拓展：scale_fill_manual()函数:

  #自定义颜色：输入的变量长度与分组变量的长度一致

  #scale_fill_manual(..., values, aesthetics = "fill", breaks = waiver(), na.value = "grey50")

###################

## 正负两极不同的颜色

csub <- subset(climate, Source == "Berkeley" & Year >= 1900)

# 思路：颜色根据正负数来填充？怎么识别正负数呢？

# 只能新建一个字段；该字段描述来正和负；

# 指定fill为该字段；

csub$pos <- csub$Anomaly10y >= 0 

csub

ggplot(csub, aes(x = Year, y = Anomaly10y, fill = pos)) +

 geom_bar(stat = "identity", position = "identity")

# 使用scale_fill_manual()修改颜色；guide = FALSE参数去掉图例

ggplot(csub, aes(x = Year, y = Anomaly10y, fill = pos)) + 

 geom_bar(stat = "identity", position = "identity", color = "black", size = 0.25) +

 scale_fill_manual(values = c("#CCEEFF", "#FFDDDD"), guide = FALSE)

  

###################

## 调整条形的宽度和间距（width参数）：

library(gcookbook)

# width默认是0.9

ggplot(pg_mean, aes(x = group, y = weight)) +

 geom_bar(stat = "identity") 

ggplot(pg_mean, aes(x = group, y = weight)) + 

 geom_bar(stat = "identity", width = 0.5)

# width最大值只能设置为1:

ggplot(pg_mean, aes(x = group, y = weight)) + 

 geom_bar(stat = "identity", width = 1)

## 调节分组条形图之间的间距：

# 默认的同一分组之间的条形是没有间距的：

ggplot(cabbage_exp, aes(x = Date, y = Weight, fill = Cultivar)) +

 geom_bar(stat = "identity", width = 0.5, position = "dodge")

# 只需要将position_dodge参数设置的比width参数大一些就好了！

ggplot(cabbage_exp, aes(x = Date, y = Weight, fill = Cultivar)) +

 geom_bar(stat = "identity", width = 0.5, position = position_dodge(0.7))

ggplot(cabbage_exp, aes(x = Date, y = Weight, fill = Cultivar)) +

 geom_bar(stat = "identity", width = 0.5, position = position_dodge(0.3))

ggplot(cabbage_exp, aes(x = Date, y = Weight, fill = Cultivar)) +

 geom_bar(stat = "identity", width = 0.5, position = position_dodge(0))

# 思考：position_dodge有什么含义？为什么比width大就会有间隙吗？

 # position_dodge是从一个柱子的右边到另一个柱子的右边的距离

 # 因为默认的position_dodge()里的内容一定是和width相等的

 

###################

## 堆积柱状图：

# position的默认值为stack；

# 即如果不设置position，并且设置了分组变量，就是画堆积图；

library(gcookbook)

ggplot(cabbage_exp, aes(x = Date, y = Weight, fill = Cultivar)) +

 geom_bar(stat = "identity")

# 修改图例堆积的顺序：guides()

ggplot(cabbage_exp, aes(x = Date, y = Weight, fill = Cultivar)) +

 geom_bar(stat = "identity") +

 guides(fill = guide_legend(reverse = T))

# 修改图形堆积顺序：修改因子水平

cabbage_exp$Cultivar <- factor(cabbage_exp$Cultivar, levels = c("c52", "c39"))

ggplot(cabbage_exp, aes(x = Date, y = Weight, fill = Cultivar)) +

 geom_bar(stat = "identity")

ggplot(cabbage_exp, aes(x = Date, y = Weight, fill = Cultivar)) +

 geom_bar(stat = "identity", color = "black") + 

 guides(fill = guide_legend(reverse = T)) +

 scale_fill_brewer(palette = "Pastel1")

###################

## 修改标签：

library(ggplot2)

library(gcookbook)

cabbage_exp <- cabbage_exp

# 标签位置的设定，vjust参数：

ggplot(cabbage_exp, aes(x = interaction(Date, Cultivar), y = Weight)) +

 geom_bar(stat = "identity") +

 geom_text(aes(label = Weight), vjust = 1.5, colour = "white")

ggplot(cabbage_exp, aes(x = interaction(Date, Cultivar), y = Weight)) + 

 geom_bar(stat = "identity") +

 geom_text(aes(label = Weight), vjust = -0.2, colour = "white")

# 为了防止标签跑出图形，可以调整y轴的范围：

# 方法一：ylim()函数：

ggplot(cabbage_exp, aes(x = interaction(Date, Cultivar), y = Weight)) +

 geom_bar(stat = "identity") +

 geom_text(aes(label = Weight), vjust = -0.2) +

 ylim(0, max(cabbage_exp$Weight) * 1.05)

# 方法二：以weight为基准，调节y值，图形高度会自动适配：

ggplot(cabbage_exp, aes(x = interaction(Date, Cultivar), y = Weight)) + 

 geom_bar(stat = "identity") +

 geom_text(aes(y = Weight + 0.1, label = Weight))

# 分组柱状图加标签：需要设定position_dodge(),以调整字体适合位置

ggplot(cabbage_exp, aes(x = interaction(Date, Cultivar), y = Weight, fill = Cultivar)) + 

 geom_bar(stat = "identity", position = "dodge") + 

 geom_text(aes(label = Weight), vjust = 1.5, color = "white", position = position_dodge(0.9), size = 3) 

 # width参数默认是0.9，position_dodge(0.9)，这样子可以让标签位于柱子中间

library(plyr)

# 堆积柱状图添加label

ce <- arrange(cabbage_exp, Date, Cultivar)

ce <- ddply(ce, "Date", transform, label_y = cumsum(Weight))

ce$Cultivar <- factor(ce$Cultivar, levels = c("c52", "c39"))

ggplot(ce, aes(x = Date, y = Weight, fill = Cultivar)) + 

 geom_bar(stat = "identity") + 

 geom_text(aes(y = label_y, label = Weight), vjust = 1.5, colour = "white")

# 修改标签至中央

ce <- arrange(cabbage_exp, Date, Cultivar)

ce <- ddply(ce, "Date", transform, label_y = cumsum(Weight) - 0.5 * Weight)

ce$Cultivar <- factor(ce$Cultivar, levels = c("c52", "c39"))

ggplot(ce, aes(x = Date, y = Weight, fill = Cultivar)) + 

 geom_bar(stat = "identity") + 

 geom_text(aes(y = label_y, label = Weight), vjust = 1.5, colour = "white")

# 添加单位，并修改颜色模式

ggplot(ce, aes(x = Date, y = Weight, fill = Cultivar)) +

 geom_bar(stat = "identity", colour = "black") +

 geom_text(aes(y = label_y, label = paste(format(Weight,nsmall = 2), "kg")), size = 4) +

 scale_fill_brewer(palette = "Pastel1")

###################

## 柱状图的拓展：克里夫兰点图绘制

library(gcookbook)

tophit <- tophitters2001[1:25,]

# 从最基本的散点图出发

ggplot(tophit, aes(x = avg, y = name)) +

 geom_point()

# reorder排序一定要熟练：前面柱状图排序讲过

ggplot(tophit, aes(x = avg, y = reorder(name, avg))) + 

 geom_point(size = 3) + #修改点的大小

 theme_bw() + #修改背景

 theme(panel.grid.major.x = element_blank(), #设置纵向网格线为空

    panel.grid.minor.x = element_blank(), #设置横向网格线为虚线（dashed）

    panel.grid.major.y = element_line(colour = "grey60", linetype = "dashed"))

# 颠倒图形的x轴和y轴

ggplot(tophit, aes(x = reorder(name, avg), y = avg))+ 

 geom_point(size = 3) + #修改点的大小

 theme_bw() + #修改背景

 theme(panel.grid.major.y = element_blank(), 

    panel.grid.minor.y = element_blank(), 

    panel.grid.major.x = element_line(colour = "grey60", linetype = "dashed"))

# 按照lg和avg对name进行排序，先按lg排，再按avg排

nameorder <- tophit$name[order(tophit$lg, tophit$avg)]

# 将name变量转化为因子，因子水平设定位nameorder：

tophit$name <- factor(tophit$name, levels = nameorder)

## 绘制彩色克利夫兰点图：

ggplot(tophit, aes(x = avg, y = name)) +

 geom_segment(aes(yend = name), xend = 0, colour = "grey50") +

 geom_point(size = 3, aes(color = lg)) + # 设置分组变量

 # limits限定颜色先后顺序：

 scale_colour_brewer(palette = "Set1", limits = c("NL", "AL")) +

 theme_bw() + 

 theme(panel.grid.major.y = element_blank(), #去除横线网格线

    legend.position = c(1, 0.55), # 设置图例的位置：这里的1指的是与x轴的比例；

    #legend.justification = c(1, 0.5) #表示图例右边缘中点；

    # (1, 0) 表示右下角，(0, 1) 表示左上角，以此类推；

    legend.justification = c(1, 0.5))

 

# 分面绘制：facet_grid()函数：

 ggplot(tophit, aes(x = avg, y = name)) +

  geom_segment(aes(yend = name), xend = 0, colour = "grey50") +

  geom_point(size = 3, aes(color = lg)) + # 设置分组变量

 # guides去除图例：

 scale_colour_brewer(palette = "Set1", limits = c("NL", "AL"), guide = F) +

  theme_bw() +

  theme(panel.grid.major.y = element_blank()) +

  # 一列多行：lg~ 行数等于lg的种类数目；

  # scales设置每个分块的单位宽度；space设置每个分块的宽度；

  facet_grid(lg~.,scales = "free_y", space = "free_y")

  

  

  

# * 第三章：各式各样的饼图 ------------------------------------------------------------

library(ggplot2)

# 技巧篇--ggplot绘制各种饼图：

mpg <- mpg

ggplot(mpg, aes(class)) +

 geom_bar()

# 把y轴方向扭曲来，柱子都边成了弯的：

ggplot(mpg, aes(class)) +

 geom_bar() +

 coord_polar(theta = "y")

# 把x轴方向扭曲来，柱子都从一个中心出发：

ggplot(mpg, aes(class)) +

 geom_bar() +

 coord_polar(theta = "x")

# 加上颜色分组：

ggplot(mpg, aes(class)) +

 geom_bar(aes(fill = drv)) + 

 coord_polar(theta = "y")

# 加上颜色分组：

ggplot(mpg, aes(class)) +

 geom_bar(aes(fill = drv)) + 

 coord_polar(theta = "x")

# 如何绘制正常的饼图？

ggplot(mpg, aes(1, fill = class)) +

 geom_bar(width = 0.5) 

ggplot(mpg, aes(1, fill = class)) +

 geom_bar(width = 0.5) + 

 coord_polar(theta = "y")

# 加上标签：

ggplot(mpg, aes(1, fill = class)) +

 geom_bar(width = 0.5) + 

 coord_polar(theta = "y") +

 geom_text(stat = "count", aes(label = scales::percent(..count../100)),

      size = 3, position = position_stack(vjust = 0.5))

# 课后作业：如何使用position_stack()修改标签位置呢？

ce <- arrange(cabbage_exp, Date, Cultivar)

ce <- ddply(ce, "Date", transform, label_y = cumsum(Weight) - 0.5 * Weight)

ce$Cultivar <- factor(ce$Cultivar, levels = c("c52", "c39"))

ggplot(ce, aes(x = Date, y = Weight, fill = Cultivar)) + 

 geom_bar(stat = "identity") + 

 geom_text(aes(y = label_y, label = Weight), vjust = 1.5, colour = "white")

# 使用position_stack()可以大大节约我们的代码：

ggplot(cabbage_exp, aes(x = Date, y = Weight, fill = Cultivar)) +

 geom_bar(stat = "identity") + 

 geom_text(aes(label = Weight), position = position_stack(vjust = 0.5), colour = "white")