R语言 使用ggplot2绘制好看的分组散点图

我们以iris数据集为例，该数据集包括花萼的长度和宽度，花瓣的长度和宽度，以及物种，如下图：

本文我们要绘制不同物种下花萼的长度和宽度的分布情况，以及二者之间的相关性关系。

1. 首先载入ggplot2包，

library(ggplot2)

2. 然后进行ggplot(data = NULL, mapping = aes(), ..., environment = parent.frame())绘制，在绘制中第一个参数是数据，第二个参数是数据映射，是绘制的全局变量，其中包含的参数有x,y,color,size,alpha,shape等。

例如：ggplot(iris, aes(Sepal.Length, Sepal.Width, color = Species))，然后通过快捷散点绘制

+geom_point(size = 2.0, shape = 16)，颜色代表不同的物种，如下图：

3. 上面显示的是最原始的散点绘制，通过颜色区分不同的物种，那么如何进行效果的提升呢？

首先是可以进行分面，使得不同物种的对比效果更为显著，这里使用+facet_wrap( ~ Species)，效果如下：

4. 通过分面后对比效果好了不少，如果想看下不同物种下花萼长度与宽度的关系呢？可以使用+geom_smooth(method = "loess")，效果图如下：

5. 通过上面的操作效果好了很多，但是还是感觉不够高大上，那我们可以使用library(ggthemes)这个包进行精修一下，通过修改theme，使用+theme_solarized()，效果如下：

还有更多的theme选择，例如+theme_wsj()，效果如下：

这样我们的图是不是高大上了很多呢，所以其实数据可视化也没有多难。最后给下源码：

library(ggthemes)
library(ggplot2)

ggplot(iris, aes(Sepal.Length, Sepal.Width, color = Species)) +
 geom_point(size = 2.0, shape = 16) +
 facet_wrap( ~ Species) +
 geom_smooth(method = "loess")+

 theme_wsj()

补充：R语言 画图神器ggplot2包

ggplot2

R语言里画图最好用的包啦。感觉图都挺清晰的，就懒得加文字了（或者以后回来补吧>.）前面几个图挺基础的，后面也许会有没见过的ggplot用法哦。

Install Package

install.packages("ggplot2")
library(ggplot2)

Scatter Plot

为了方便展示，用gapminder的数据

if(!require(gapminder)) install.packages("gapminder"）
 library(gapminder)
gapminder

数据大概是这样的

假设我们现在想要知道2007年lifeExp和人均GDP之间的关系。

先筛选数据

library(dplyr)
gapminder_2007 <- gapminder %>%
 filter(year == 2007)

画lifeExp和gdpPercap关系的散点图，x为gdpPercap，y为lifeExp。

ggplot(gapminder_2007,aes(x = gdpPercap, y = lifeExp))+geom_point()

看的出来lifeExp与gdpPercap存在近似lifeExp=log(gdpPercap)的关系，对x轴的数值进行log值处理。另外，为了呈现更多信息，用颜色标记国家所在的洲，并用点的大小表示人口数量。

ggplot(gapminder_2007,aes(x = gdpPercap, y = lifeExp, color = continent, size = pop))+
 geom_point()+scale_x_log10()+theme_minimal()+
 labs(x = "GDP per capita",
 y = "Life expectancy",
 title = "Life expectancy increases as GDP per capita increases",
 caption = "Data source: gapminder")

另外一种呈现方式如下：

加入了回归线和坐标轴的histogram。

plot <- ggplot(gapminder_2007, aes(x = gdpPercap, y = lifeExp)) + 
 geom_point()+geom_smooth(method="lm")+scale_x_log10()+
 labs(x = "GDP per capita",
 y = "Life expectancy",
 title = "Life expectancy increases as GDP per capita increases",
 caption = "Data source: gapminder")
ggMarginal(plot, type = "histogram", fill="transparent")
#ggMarginal(plot, type = "boxplot", fill="transparent")

Histogram

gapminder_gdp2007 <- gapminder %>%
 filter(year == 2007, continent == "Americas") %>%
 mutate(country = fct_reorder(country,gdpPercap,last))
ggplot(gapminder_gdp2007, aes(x=country, y = gdpPercap))+
 geom_col(fill="skyblue", color="black")+
 labs(x = "Country",
 y = "GDP per capita",
 title = "GDP per capita in North America and South America, 2007",
 caption = "Data source: gapminder")+
 coord_flip()+theme_minimal()

Line Plot

gapminder_pop <- gapminder %>%
 filter(country %in% c("United States","China"))
ggplot(gapminder_pop,aes(x = year, y = pop, color = country))+
 geom_line(lwd = 0.8)+theme_light()+
 labs(x = "Year",
 y = "Population",
 title = "Population in China and United States, 1953-2007",
 caption = "Data source: gapminder")

Facet Plot

gapminder_gdp <- gapminder %>%
 group_by(year, continent) %>%
 summarize(avg_gdp = mean(gdpPercap))
ggplot(gapminder_gdp,aes(x = year, y = avg_gdp, color = continent))+
 geom_line(lwd = 0.8)+theme_light()+facet_wrap(~continent)+
 labs(x = "Year",
 y = "Average GDP per capita",
 title = "Average GDP per capita change in different continent",
 caption = "Data source: gapminder")+
 scale_x_continuous(breaks=c(1955,1970,1985,2000))

Path Plot

gapminder_lifeexp <- gapminder %>%
 filter(year %in% c(1957,2007), continent == "Europe") %>%
 arrange(year) %>%
 mutate(country = fct_reorder(country,lifeExp,last))
ggplot(gapminder_lifeexp) +geom_path(aes(x = lifeExp, y = country),
 arrow = arrow(length = unit(1.5, "mm"), type = "closed")) +
 geom_text(
 aes(x = lifeExp,
 y = country,
 label = round(lifeExp, 1),
 hjust = ifelse(year == 2007,-0.2,1.2)),
 size =3,
 family = "Bookman",
 color = "gray25")+
 scale_x_continuous(limits=c(45, 85))+
 labs(
 x = "Life expectancy",
 y = "Country",
 title = "People live longer in 2007 compared to 1957",
 subtitle = "Life expectancy in European countries",
 caption = "Data source: gapminder"
 )
 

Density Plot

gapminder_1992 <- gapminder %>%
 filter(year == 1992)
ggplot(gapminder_1992, aes(lifeExp))+theme_classic()+
 geom_density(aes(fill=factor(continent)), alpha=0.8) + 
 labs(
 x="Life expectancy",
 title="Life expectancy group by continent, 1992", 
 caption="Data source: gapminder",
 fill="Continent")

Slope Chart

gapminder_lifeexp2 <- gapminder %>%
 filter(year %in% c(1977,1987,1997,2007),
 country %in% c("Canada", "United States","Mexico","Haiti","El Salvador",
  "Guatemala","Jamaica")) %>%
 mutate(lifeExp = round(lifeExp))
ylabs <- subset(gapminder_lifeexp2, year==head(year,1))$country
yvals <- subset(gapminder_lifeexp2, year==head(year,1))$lifeExp
ggplot(gapminder_lifeexp2, aes(x=as.factor(year),y=lifeExp)) +
 geom_line(aes(group=country),colour="grey80") +
 geom_point(colour="white",size=8) +
 geom_text(aes(label=lifeExp), size=3, color = "black") +
 scale_y_continuous(name="", breaks=yvals, labels=ylabs)+
 theme_classic()+
 labs(title="Life Expectancy of some North America countries change from 1977 to 2007") + 
 theme(axis.title=element_blank(),
 axis.ticks = element_blank(),
 plot.title = element_text(hjust=0.5))

以上为个人经验，希望能给大家一个参考，也希望大家多多支持猪先飞。如有错误或未考虑完全的地方，望不吝赐教。