R 数据分析学习笔记

模拟成绩
用runif 和rnorm
rnorm(n, mean = 0, sd = 1)
n 为产生随机值个数（长度），mean 是平均数， sd 是标准差 。
使用该函数的时候后，一般要赋予它 3个值.

=seq(10378001,10378100)
> num
  [1] 10378001 10378002 10378003 10378004 10378005 10378006
  [7] 10378007 10378008 10378009 10378010 10378011 10378012
 [13] 10378013 10378014 10378015 10378016 10378017 10378018
 [19] 10378019 10378020 10378021 10378022 10378023 10378024
 [25] 10378025 10378026 10378027 10378028 10378029 10378030
 [31] 10378031 10378032 10378033 10378034 10378035 10378036
 [37] 10378037 10378038 10378039 10378040 10378041 10378042
 [43] 10378043 10378044 10378045 10378046 10378047 10378048
 [49] 10378049 10378050 10378051 10378052 10378
> x1=round(runif(100,min=80,max=100))
> x1
  [1]  95  83  95  87  97  89  87  85  87  91  94  82  98
 [14]  93  97  84  83  94 100  96  88  94  98  96  83  88
 [27]  91  84  91  86  98  95  86  99  97  86  91  96 100
 [40]  88  99  95  87  93  95  81  86  84  82  88  89  96
 [53]  81  88  97  82  86  87  93  89  80  86  88  86  96
 [66]  97 100  90 100  82  88 100  82  87  93  85  91  81
 [79]  87  91  93  96  96  98  98  92  99  92  92  86  91
 [92]  92  80  97  91  86  97  81  83  96
> x2=round(rnorm(100,mean=80,sd=7))
> x2
  [1]  77  79  72  74  75  80  92  83  82  77  74  78  87
 [14]  83  80  87  88  74  75  70  81  75  81  74  83 100
 [27]  78  78  75  78  74  78  72  86  81  74  70  79  80
 [40]  66  89  72  85  88  79  86  84  77  86  81  87  73
 [53]  86  88  86  84  78  88  83  79  83  66  67  79  82
 [66]  83  92  76  84  78  83  77  80  79  73  85  88  83
 [79]  81  74  72  77  69  85  85  82  67  76  87  78  75
 [92]  80  88  68  96  76  89  72  78  73
> 
> x3=round(rnorm(100,mean=80,sd=18))
> x3
  [1]  49  74  83  97  70  72  80  74  82  94  76 129 122
 [14]  65  83  72  78  91  71  43 106  62  93  71 109  76
 [27] 102 107  75 101  87  86  76  98  86  73  79  88  56
 [40]  77  67 102  82  96  95  63  82 112  80  52  27  84
 [53]  91  73  44  70  70  60  91  86 112 101  71 102  94
 [66]  84  41  90  79  86  96  81  70  85  87  62  87  49
 [79] 113  65 113  75  25  53  54  88  64  96  73  53  94
 [92]  97  83  55  38 110  92  85  91  80
> x3[which(x3>100)]=100
> x3
  [1]  49  74  83  97  70  72  80  74  82  94  76 100 100
 [14]  65  83  72  78  91  71  43 100  62  93  71 100  76
 [27] 100 100  75 100  87  86  76  98  86  73  79  88  56
 [40]  77  67 100  82  96  95  63  82 100  80  52  27  84
 [53]  91  73  44  70  70  60  91  86 100 100  71 100  94
 [66]  84  41  90  79  86  96  81  70  85  87  62  87  49
 [79] 100  65 100  75  25  53  54  88  64  96  73  53  94
 [92]  97  83  55  38 100  92  85  91  80

> mean(x)
[1] NA
警告信息：
In mean.default(x) : 参数不是数值也不是逻辑值：回覆NA
> colMeans(x)
        num          x1          x2          x3 
10378050.50       90.65       79.98       83.55 
       x1.1        x2.1        x3.1 
      90.56       79.75       78.73 
> colMeans(x)[c("x1","x2","x3")]
   x1    x2    x3 
90.65 79.98 83.55 
> apply(x,2,mean)
        num          x1          x2          x3 
10378050.50       90.65       79.98       83.55 
       x1.1        x2.1        x3.1 
      90.56       79.75       78.73 
> 
> num=seq(10378001,10378100)
> num
  [1] 10378001 10378002 10378003 10378004 10378005 10378006
  [7] 10378007 10378008 10378009 10378010 10378011 10378012
 [13] 10378013 10378014 10378015 10378016 10378017 10378018
 [19] 10378019 10378020 10378021 10378022 10378023 10378024
 [25] 10378025 10378026 10378027 10378028 10378029 10378030
 [31] 10378031 10378032 10378033 10378034 10378035 10378036
 [37] 10378037 10378038 10378039 103
> 
> x=data.frame(x,x1,x2,x3)
> x
         num  x1 x2  x3 x1.1 x2.1 x3.1
1   10378001  98 85 117   95   77   49
2   10378002  86 84  70   83   79   74
3   10378003  92 77 105   95   72   83
4   10378004  92 78 118   87   74   97
5   10378005  89 80  64   97   75   70
6   10378006  82 92  91   89   80   72
7   10378007  91 92  77   87   92   80
8   10378008  92 93  69   85   83   74
> mean(x)
[1] NA
警告信息：
In mean.default(x) : 参数不是数值也不是逻辑值：回覆NA
> colMeans(x)
        num          x1          x2          x3 
10378050.50       90.65       79.98       83.55 
       x1.1        x2.1        x3.1 
      90.56       79.75       78.73 
> colMeans(x)[c("x1","x2","x3")]
   x1    x2    x3 
90.65 79.98 83.55 
> apply(x,2,mean)
        num          x1          x2          x3 
10378050.50       90.65       79.98       83.55 
       x1.1        x2.1        x3.1 
      90.56       79.75       78.73 
> 
> 
> apply(x,2,max)
     num       x1       x2       x3     x1.1     x2.1 
10378100      100       97      132      100      100 
    x3.1 
     100 
> apply(x,2,min)
     num       x1       x2       x3     x1.1     x2.1 
10378001       80       65       36       80       66 
    x3.1 
      25 
> apply(x,2,sum)
       num         x1         x2         x3       x1.1 
1037805050       9065       7998       8355       9056 
      x2.1       x3.1 
      7975       7873 
> apply(x,2,sum)[c("x1","x2","x3")]
  x1   x2   x3 
9065 7998 8355 
> 
> apply(x[c("x1","x2","x3")],1,sum)
  [1] 300 240 274 288 233 265 260 254 228 262 239 206 247
 [14] 234 260 266 254 278 245 282 282 259 274 277 231 216
 [27] 211 261 301 278 275 247 248 287 249 245 281 265 250
 [40] 234 275 259 232 281 252 257 226 249 288 264 247 262
 [53] 255 254 265 229 272 226 223 246 299 217 259 250 240
 [66] 251 279 236 263 273 277 255 287 248 244 249 228 223
 [79] 284 273 278 252 268 229 257 233 189 253 233 246 253
 [92] 230 244 252 258 207 238 269 256 290
> 
> which.max(apply(x[c("x1","x2","x3")],1,sum))
[1] 29

> x$num[which.max(apply(x[c("x1","x2","x3")],1,sum))]
[1] 10378029

> x$num[which.max(apply(x[c("x1","x2","x3")],1,sum))]
[1] 10378029
> hist(x$x1)
> plot(x1,x2)
> plot(x$x1,x$x2)

列联表分析
列联函数table(),柱状图绘制函数barplot()

 table(x$x1)

 80  81  82  83  84  85  86  87  88  89  90  91  92  93  94  95  96  97  98 
  3   6   4   4   6   3   6   5   3   4   4   4   8   4   3   2   5   4   5 
 99 100 
 13   4 
> barplot(table(x$x1))

* 饼图 饼图绘制函数pie() *

> pie(table(x$x1))

* 箱尾图*

1. 箱子的上下横线为样本的25%和75%分位数
2. 箱子中间的横线为样本的中位数
   3.上下延伸的直线称为尾线，尾线的尽头为最高值和最低值
3. 异常值

boxplot(x$x1,x$x2,x$x3)

* 箱线图*

 >boxplot(x[2:4],col=c("red","green","blue"),notch=T)

星像图

1. 每个观测单位的数值表示为一个图形
2. 每个图的每个角表示一个变量，字符串类型会标注在图的下方
3. 角线的长度表示表达值的大小

> stars(x[c("x1","x2","x3")])

> stars(x[c("x1","x2","x3")],full=T,draw.segments=T)

> stars(x[c("x1","x2","x3")],full=F,draw.segments=T)
> 

脸谱图

• 用五官的宽度和高度来描述数值
• 人对脸谱高度敏感和强记忆
• 适合较少样本的情况

QQ图
- 可用于判断是否是正态分布
- 直线的斜率是标准差，截距是均值
- 点的散布越接近直线，则越接近正态分布

> qqnorm(x1)

> qqline(x1)

> qqnorm(x3)