R 数据可视化

class: center, middle, inverse, title-slide

# R 数据可视化
## 基础作图
### Chris Qi
### 2018

---

background-image: url(https://www.dropbox.com/s/zk8948u32y01sj8/logo.png?dl=1)
background-size: 700px
background-position: 50% 50%

---
background-image: url(https://www.dropbox.com/s/zk8948u32y01sj8/logo.png?dl=1)
background-size: 200px
background-position: 95% 3%

---
# R 基础作图快速入门

导语：用数据说话的今天，随着数据量的不断增加，数据可视化成为将数字变成可用的信息的一个重要方式。

--
* 基本绘图系统（Base Plotting System）

--
* Lattice绘图系统（Lattice Plotting System）

--
* ggplot2 绘图系统（ggplot2 Plotting System）

---
# 常用图形

* 直方图 histogram

* 散点图 scatter

* 箱图 boxplot

---
# 基本绘图系统

* 人人都是艺术家：绘图开始于空白，一笔一划添加内容

* 事先计划好，直观地反映绘图和分析数据的逻辑

* 用基本R绘图需要走两步
  * 图
  * 修饰/添加=执行一系列函数
  
* quick, easy, dirty

---
# 直方图（频数）

```r
library(MASS)
hist(Cars93$Horsepower,main="hist() plot")
```

---
# 直方图（频数）

![](slides_basicGraphs_files/figure-html/unnamed-chunk-2-1.png)

---

# 直方图 histogram（频率）
'truehist()'频率直方图的函数，来自MASS这个包：

```r
library(MASS)
truehist(Cars93$Horsepower,main="hist() plot")
```

---
# 直方图 histogram（频率）
'truehist()'频率直方图的函数，来自MASS这个包：

```r
library(MASS)
truehist(Cars93$Horsepower,main="hist() plot")
```

![](slides_basicGraphs_files/figure-html/unnamed-chunk-4-1.png)

---
# 散点图 scatter

```r
library(MASS)
plot(UScereal$sugars, UScereal$calories)
title("plot(sugars and calories in US cereals)")
```

---
# 散点图 scatter 
![](slides_basicGraphs_files/figure-html/unnamed-chunk-6-1.png)

---
散点图中的菊花图，让你更懂数据

```r
# Set up a side-by-side plot array
par(mfrow=c(1,2))

# Create the standard scatterplot
plot(rad~zn,data=Boston)

# Add the title
title("Standard scatterplot")

# Create the sunflowerplot
sunflowerplot(rad~zn,data=Boston)

# Add the title
title("Sunflower plot")
```

---
散点图中的菊花图，让你更懂数据

![](slides_basicGraphs_files/figure-html/unnamed-chunk-8-1.png)

---
# 箱图 boxplot

```r
boxplot(Temp~Month, data=airquality)
```

![](slides_basicGraphs_files/figure-html/unnamed-chunk-9-1.png)

---
# Lattice graphics 绘图系统:
* 需要安装 install.packages("lattice")
  * 有这些函数可以调用：xyplot/bwplot/histogram/stripplot/dotplot/splom/levelplot.contourplot
  
* 格式xyplot(y~x|f*g,data)，f*g是分类变量

* 特别适用于观测变量间的交互：在变量z的不同水平，变量y如何随着x变化

---
lattice 绘图

```r
library(MASS)
library(lattice) #需要先安装install.packages("lattice")
xyplot(Temp~Ozone, data=airquality) 
```

![](slides_basicGraphs_files/figure-html/unnamed-chunk-10-1.png)

---
lattice 绘图，考察交互关系

```r
airquality$Month<-factor(airquality$Month)
xyplot(Temp~Ozone|Month, data=airquality, layout=c(5,1))
```

![](slides_basicGraphs_files/figure-html/unnamed-chunk-11-1.png)

---
# ggplot2 绘图系统:

```r
library(MASS)
library(ggplot2)
title <-
 "ggplot2 plot of \n UScereal$calories vs. \n UScereal$sugars"
basePlot <- ggplot(UScereal, aes(x = sugars, y = calories))
basePlot +
 geom_point(shape = as.character(UScereal$shelf), size = 3) +
 annotate("text", label = title, x = 3, y = 400,
 colour = "red")
```

---
ggplot2:
![](slides_basicGraphs_files/figure-html/unnamed-chunk-13-1.png)

---
# 探索性数据可视化:

目的：

* 了解数据特征，找到数据中的模式，形成分析策略

* 图与数字相互验证，帮助发现错误，用于交流结果

---
# 探索性数据可视化:

特点：

* 快速，通常呈现在屏幕设备上

* 不需要过分注重图是否漂亮

---
探索小鸡的成长之路：ChickWeight数据集

```r
plot(ChickWeight)
```

![](slides_basicGraphs_files/figure-html/unnamed-chunk-14-1.png)

---
# 解释（分析）性数据可视化：

--
* 整合文字，数字， 图， 表等来讲一个美丽的故事

--
* 用多种方式显示数据的特征

--
* 不要让工具主宰分析
  * 研究问题的重要性 > 作图漂亮
  
  
--
* 使用适当的图标，尺度等
  * 完备性，一图胜千言

让别人明白数据，便于交流信息，呈现你的数据发现

---
我们用一个简单的冬季取暖数据来建构一个简单的分析图：

--
* Temp: 一周的室外气温

--
* Gas: 该周的天然气用量

--
* Insul: 有两个值的分类变量，分别代表房子升级改造前后

---
简单绘图看一下：

```r
# Plot whiteside data
plot(whiteside)
```

![](slides_basicGraphs_files/figure-html/unnamed-chunk-15-1.png)

---
简单的探索性散点图：
不要完全依赖默认设置，适当修改，让图更有意义容易理解
做一个气温与天然气用量的散点图，加上轴标签：

```r
plot(whiteside$Temp,whiteside$Gas,
     xlab="Outside temperature",
     ylab= "Heating gas consumption")
```

---
简单的探索性散点图：
不要完全依赖默认设置，适当修改，让图更有意义容易理解
做一个气温与天然气用量的散点图，加上轴标签：

![](slides_basicGraphs_files/figure-html/unnamed-chunk-17-1.png)

---
'plot()'是一个一般性的函数，会根据数据类型的不同而作出不同的图，例如下面，我们把它用在'whiteside$Insul'上，得到一个完全不一样的图

```r
plot(whiteside$Insul)
```

![](slides_basicGraphs_files/figure-html/unnamed-chunk-18-1.png)

---
使用形状，颜色和参考线来增加细节，我们用Cars93这个数据集来做一个高级一点的图来展示三个变量间的关系：

--
* Price: 某款车的平均售价

--
* Max.Price: 该款车的最高售价

--
* Min.Price: 该款车的最低售价

---
使用 'names()' 来看看这个Cars93 都有什么变量

```r
names(Cars93)
```

```
##  [1] "Manufacturer"       "Model"              "Type"              
##  [4] "Min.Price"          "Price"              "Max.Price"         
##  [7] "MPG.city"           "MPG.highway"        "AirBags"           
## [10] "DriveTrain"         "Cylinders"          "EngineSize"        
## [13] "Horsepower"         "RPM"                "Rev.per.mile"      
## [16] "Man.trans.avail"    "Fuel.tank.capacity" "Passengers"        
## [19] "Length"             "Wheelbase"          "Width"             
## [22] "Turn.circle"        "Rear.seat.room"     "Luggage.room"      
## [25] "Weight"             "Origin"             "Make"
```

---
最高价vs.平均价，红色三角

```r
plot(Cars93$Max.Price, Cars93$Price,col="red",pch=17)
```

![](slides_basicGraphs_files/figure-html/unnamed-chunk-20-1.png)

---
添加最高价vs. 平均价，蓝色，圆点

```r
plot(Cars93$Max.Price, Cars93$Price,col="red",pch=17)
points(Cars93$Min.Price,Cars93$Price,col="blue",pch=16)
```

![](slides_basicGraphs_files/figure-html/unnamed-chunk-21-1.png)

---
添加一条参考线，0为截距，1为斜率，虚线

```r
plot(Cars93$Max.Price, Cars93$Price,col="red",pch=17)
points(Cars93$Min.Price,Cars93$Price,col="blue",pch=16)
abline(a = 0, b = 1, lty = 2)
```

![](slides_basicGraphs_files/figure-html/unnamed-chunk-22-1.png)

---
'with()' 更简便的绘图函数，再不用每个变量都引用数据集。

```r
with(airquality, plot(Wind~Temp))
title(main="wind and temp in NYC") #使用title进行修饰
```

![](slides_basicGraphs_files/figure-html/unnamed-chunk-23-1.png)

---

```r
#另外一种方式添加标题，一样的结果
with(airquality, plot(Wind, Temp,
     main="wind and temp in NYC")) 
```

![](slides_basicGraphs_files/figure-html/unnamed-chunk-24-1.png)

---

```r
#有用的 type="n"
with(airquality, plot(Wind, Temp,
                      main="Wind and Temp in NYC",
                      type="n"))
```

![](slides_basicGraphs_files/figure-html/unnamed-chunk-25-1.png)

---

```r
with(airquality, plot(Wind, Temp,
                      main="Wind and Temp in NYC",
                      type="n"))
with(subset(airquality, Month==9), 
     points(Wind, Temp, col="red"))
```

![](slides_basicGraphs_files/figure-html/unnamed-chunk-26-1.png)

---

```r
with(airquality, plot(Wind, Temp,
                      main="Wind and Temp in NYC",
                      type="n"))
with(subset(airquality, Month==9), 
     points(Wind, Temp, col="red"))
with(subset(airquality, Month==8), 
     points(Wind, Temp, col="blue"))
```

![](slides_basicGraphs_files/figure-html/unnamed-chunk-27-1.png)

---

```r
with(airquality, plot(Wind, Temp,
                      main="Wind and Temp in NYC",
                      type="n"))
with(subset(airquality, Month==9), 
     points(Wind, Temp, col="red"))

with(subset(airquality, Month==8), 
     points(Wind, Temp, col="blue"))

with(subset(airquality, Month==7), 
     points(Wind, Temp, col="green"))
#完成了用不同的颜色表示不同的月份的点
```

---
![](slides_basicGraphs_files/figure-html/unnamed-chunk-29-1.png)

---

```r
with(airquality, plot(Wind, Temp,
                      main="Wind and Temp in NYC",
                      type="n"))
with(subset(airquality, Month==9), 
     points(Wind, Temp, col="red"))

with(subset(airquality, Month==8), 
     points(Wind, Temp, col="blue"))

with(subset(airquality, Month==7), 
     points(Wind, Temp, col="green"))

with(subset(airquality, Month %in% c(5,6)), 
     points(Wind, Temp, col="black"))
```

---
![](slides_basicGraphs_files/figure-html/unnamed-chunk-31-1.png)

---

```r
with(airquality, plot(Wind, Temp,
                      main="Wind and Temp in NYC",
                      type="n"))
with(subset(airquality, Month==9), 
     points(Wind, Temp, col="red"))

with(subset(airquality, Month==8), 
     points(Wind, Temp, col="blue"))

with(subset(airquality, Month==7), 
     points(Wind, Temp, col="green"))

with(subset(airquality, Month %in% c(5,6)), 
     points(Wind, Temp, col="black"))

fit<-lm(Temp~Wind, airquality) #拟合一个回归模型
abline(fit,lwd=2) #把回归线加进去

#加入图例，说明颜色含义
legend("topright", pch=1, #右上角，空心圆点
       col=c("red","blue","green","black"),
       legend=c("Sep","August","July","Other"))
#使用基本函数绘图，需要事先计划，运行很多函数，
#可以直观的反映我们绘图的逻辑
#图例要仔细一一对应
```

---
![](slides_basicGraphs_files/figure-html/unnamed-chunk-33-1.png)

---

完结撒花

.center[![](https://www.dropbox.com/s/9w3nxj755iqxnrz/sahua.gif?dl=1)]