DatistEQ之ggplot2单变量绘图

本文主要讲述使用ggplot2组件，进行单变量的绘图。

构造一个随机数据集。

#构造数据集，随机创建两列数据
set.seed(1234)
wdata <- data.frame(
  sex=factor(rep(c("F", "M"), each=200)),
  weight=c(rnorm(200, 55), rnorm(200, 58))
)
#wdata等价于前节点推送进来的数据框inputtable

前10行数据

先绘制一个图层a,后面逐步添加图层

a <- ggplot(wdata, aes(x=weight))

可能添加的图层有：

对于一个连续变量：

• 面积图：geom_area()
• 密度图：geom_density()
• 点图：geom_dotplot()
• 频率多边图：geom_freqpoly()
• 直方图：geom_histogram()
• 经验累积密度图：stat_ecdf()
• QQ图：stat_qq()

对于一个离散变量：

• 条形图：geom_bar()

单变量绘图

一、单变量：连续型

1. 面积图

a+geom_area(stat = "bin")
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")

面积图

改变颜色

a+geom_area(aes(fill=sex), stat = "bin", alpha=0.6)+ theme_classic()
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")

带颜色的面积图

注意：y轴默认为变量weight的数量即count，如果y轴要显示密度，可用以下代码：

a+geom_area(aes(y=..density..), stat = "bin")
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")

密度面积图

可以通过修改不同属性如透明度、填充颜色、大小、线型等自定义图形：

2. 密度图

使用以下函数：

• geom_density()：绘制密度图
• geom_vline()：添加竖直线
• scale_color_manual()：手动修改颜色

a <- ggplot(wdata, aes(x=weight)) + geom_density()
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")

密度图

根据sex修改颜色，将sex映射给line颜色

a <- ggplot(wdata, aes(x=weight)) 
a+geom_density(aes(color=sex))
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")

带颜色的密度图

修改填充颜色以及透明度

a+geom_density(aes(fill=sex), alpha=0.4)
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")

带颜色填充的密度图

添加均值线以及手动修改颜色

if (!require("plyr")) install.packages("plyr")
library(plyr)
mu <- ddply(wdata, "sex", summarise, grp.mean=mean(weight))

a+geom_density(aes(color=sex))+
  geom_vline(data=mu, aes(xintercept=grp.mean, color=sex), linetype="dashed")+
  scale_color_manual(values = c("red", "blue"))
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")

带均值线的密度图

3. 点图

a+geom_dotplot()
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")

点图

将sex映射给颜色

a+geom_dotplot(aes(fill=sex))
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")

带颜色的点图

手动修改颜色

a+geom_dotplot(aes(fill=sex))+
  scale_fill_manual(values=c("#999999", "#E69F00"))
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")

修改点图的颜色

4. 频率多边图

a+geom_freqpoly()
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")

频率多边图

y轴显示为密度

a+geom_freqpoly(aes(y=..density..)) + theme_minimal()
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")

频率多边图

修改颜色以及线型

a+geom_freqpoly(aes(color=sex, linetype=sex))+theme_minimal()
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")

频率多边图

5. 直方图

header1("直方图")
a+geom_histogram()  
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")

直方图

将sex映射给线颜色

a+geom_histogram(aes(color=sex), fill="white", position = "dodge")+theme_classic()
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")

直方图

6. 经验累积密度图

a+stat_ecdf()
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")

经验累积密度图

7. QQ图

ggplot(data = mtcars, aes(sample=mpg))+stat_qq()
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")

QQ图

二、单变量：离散型

data(mpg)
b <- ggplot(mpg, aes(x=fl))
b+geom_bar()
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")

条形图

修改填充颜色

b+geom_bar(fill="steelblue", color="black")+theme_classic()
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")

带颜色的条形图

本文中涉及的所有代码：

#构造数据集
set.seed(1234)
wdata <- data.frame(
  sex=factor(rep(c("F", "M"), each=200)),
  weight=c(rnorm(200, 55), rnorm(200, 58))
)

output(wdata)

#一个变量：连续型
library(ggplot2)
a <- ggplot(wdata, aes(x=weight)) + geom_density()
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")

#面积图
header1("面积图")
a+geom_area(stat = "bin")
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")

a+geom_area(aes(fill=sex), stat = "bin", alpha=0.6)+ theme_classic()
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")

a+geom_area(aes(y=..density..), stat = "bin")
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")

#密度图
header1("密度图")
a <- ggplot(wdata, aes(x=weight)) + geom_density()
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")

#根据sex修改颜色，将sex映射给line颜色
a <- ggplot(wdata, aes(x=weight)) 
a+geom_density(aes(color=sex))
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")

#修改填充颜色以及透明度
a+geom_density(aes(fill=sex), alpha=0.4)
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")

#添加均值线以及手动修改颜色
if (!require("plyr")) install.packages("plyr")
library(plyr)
mu <- ddply(wdata, "sex", summarise, grp.mean=mean(weight))

a+geom_density(aes(color=sex))+
  geom_vline(data=mu, aes(xintercept=grp.mean, color=sex), linetype="dashed")+
  scale_color_manual(values = c("red", "blue"))
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")

#点图
header1("点图")
a+geom_dotplot()
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")

a+geom_dotplot(aes(fill=sex))
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")

#手动修改颜色
a+geom_dotplot(aes(fill=sex))+
  scale_fill_manual(values=c("#999999", "#E69F00"))
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")

#频率多边图
header1("频率多边图")
a+geom_freqpoly()
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")

#y轴显示为密度
a+geom_freqpoly(aes(y=..density..))+
  theme_minimal()
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")

#修改颜色以及线型
a+geom_freqpoly(aes(color=sex, linetype=sex))+
  theme_minimal()
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")

header1("直方图")
a+geom_histogram()  
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")

#将sex映射给线颜色
a+geom_histogram(aes(color=sex), fill="white", position = "dodge")+theme_classic()
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")

header1("经验累积密度图")
#经验累积密度图
a+stat_ecdf()
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")

header1("QQ图")
ggplot(data = mtcars, aes(sample=mpg))+stat_qq()
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")

header1("一个离散变量")
#加载数据集
data(mpg)
b <- ggplot(mpg, aes(x=fl))
b+geom_bar()
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")

#修改填充颜色
b+geom_bar(fill="steelblue", color="black")+theme_classic()
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")

header1("两个变量：x,y皆连续")
header2("散点图")
b <- ggplot(data = mtcars, aes(x=wt, y=mpg))
b+geom_point()
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")

#将变量cyl映射给点的颜色和形状
b + geom_point(aes(color = factor(cyl), shape = factor(cyl)))
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")

#自定义颜色
b+geom_point(aes(color=factor(cyl), shape=factor(cyl)))+
  scale_color_manual(values=c("#999999", "#E69F00", "#56B4E9"))+theme_classic()
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")
  
header2("平滑线")
b+geom_smooth()
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")

#散点图+回归线
b+geom_point()+
  geom_smooth(method = "lm", se=FALSE)#去掉置信区间
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")

#使用loess方法
b+geom_point()+
  geom_smooth(method = "loess")
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")  
  
#将变量映射给颜色和形状
b+geom_point(aes(color=factor(cyl), shape=factor(cyl)))+
  geom_smooth(aes(color=factor(cyl), shape=factor(cyl)), method = "lm", se=FALSE, fullrange=TRUE)
ggsave(gettempfile(), width = 15, height = 10, dpi=300,units = "cm")  

参考文献
1、https://blog.csdn.net/woodcorpse/article/details/106552735