4-2 R语言函数 apply

#apply函数，沿着数组的某一维度处理数据
#例如将函数用于矩阵的行或列
#与for/while循环的效率相似，但只用一句话可以完成
#apply(参数)：apply(数组,维度,函数/函数名)


> x <- matrix(1:16,4,4)
> x
     [,1] [,2] [,3] [,4]
[1,]    1    5    9   13
[2,]    2    6   10   14
[3,]    3    7   11   15
[4,]    4    8   12   16

> apply(x,2,mean) #沿着x的第二维度求平均，x有两个维度，行和列，第二个维度就是沿着x的列求平均
[1]  2.5  6.5 10.5 14.5

> apply(x,2,sum) #沿着x的第二维度求和
[1] 10 26 42 58

> apply(x,1,sum)
[1] 28 32 36 40

> apply(x,1,mean)
[1]  7  8  9 10

> rowSums(x) #行的总和
[1] 28 32 36 40

> rowMeans(x) #行的平均值
[1]  7  8  9 10

> colSums(x) #列的总和
[1] 10 26 42 58

> colMeans(x) #列的平均值
[1]  2.5  6.5 10.5 14.5

> x <- matrix(rnorm(100),10,10) #随机从正态分布中取100个数据
> x
            [,1]       [,2]        [,3]       [,4]        [,5]       [,6]       [,7]
 [1,] -0.6028508  1.4642242  0.04427663  0.2871729  0.04981660 -0.8558895  0.5130530
 [2,] -1.9378240  0.2039535 -0.19909385 -0.4309858  0.85004373  0.4976094 -0.5580487
 [3,]  1.2487024  0.3279828 -0.61134011 -0.1575374 -0.29225789  0.3887533  0.3905769
 [4,] -2.5628573  0.4519969 -0.31849107 -1.4633238  0.46414326  0.3366307 -2.1061818
 [5,] -0.2568173 -0.7387934 -0.65190045 -1.5211132 -0.68554516  0.3329140 -1.3744196
 [6,] -0.3072326 -1.2575338  0.42412478 -1.3476506 -0.21221874  0.7673182 -0.4560506
 [7,]  0.1561480  0.3020903  0.36489259 -0.2507313  1.35735729 -0.2610940  0.5355151
 [8,]  0.6536334  0.3717443 -0.77679094  1.0801878  0.07262787 -0.5006976 -2.6058038
 [9,]  1.4417755 -1.2989872  1.04908993  0.5010024 -0.41921218  2.2141514  0.3646026
[10,] -1.6978768 -0.9097784  0.01689380  0.6535433  1.55588778  0.4550700  2.5595517
             [,8]       [,9]       [,10]
 [1,] -0.57296509  0.1170718 -1.89788063
 [2,]  0.06360181  1.3552013  0.83369280
 [3,] -0.44550756  0.3857978  0.24664750
 [4,]  0.51678695  0.2522804 -0.77862862
 [5,]  0.35021885 -0.2767039 -0.37358325
 [6,] -0.12660675 -1.4168734  0.86864076
 [7,]  0.69927317  0.6202195 -2.31017158
 [8,]  1.43228754  1.3257759  0.59362053
 [9,] -1.63696656  0.3467712  0.72186091
[10,] -1.02416667 -1.7024939  0.03971799


#解释：
#x赋值函数中的2*3*4分别对应行*列*组（相对应的维度即为1*2*3
#apply(x,c(1,2),mean)中1,2对应的维度为行*列，不需要考虑组，所以对每组相同位置的所有元素相加后求平均，因此输出的结果为2行3列的矩阵
#同理，apply(x,c(1,3),mean)中1,3对应的维度为行*组，所以分别对每组中的行求平均，因此输出的结果为2行4列的矩阵(x中有4个组，每组中有2行)
#同理，(2,3)就代表列*组了~

> apply(x,1,quantile,probs=c(0.25,0.75)) #quantile求数据的百分位点，可通过probs=c()进行分配
          [,1]       [,2]       [,3]       [,4]       [,5]       [,6]       [,7]
25% -0.5953794 -0.3730128 -0.2585778 -1.2921500 -0.7254813 -1.0571630 -0.1490114
75%  0.2446476  0.7496719  0.3880144  0.4231553 -0.2617889  0.2864419  0.5990434
          [,8]       [,9]      [,10]
25% -0.3573663 -0.2277163 -0.9955696
75%  0.9735492  0.9672827  0.6039250

> x <- array(rnorm(2*3*4),c(2,3,4))#表示随机从正太分布中抽取出来的24个数据，按照三维排列出来。
> x
, , 1

           [,1]      [,2]       [,3]
[1,] -0.6055074 0.1428984 -0.9020732
[2,] -0.6947868 1.3597884  0.8797562

, , 2

           [,1]       [,2]       [,3]
[1,] -0.3114873 -2.3184400  0.4499677
[2,]  0.1497819  0.1295499 -1.6927436

, , 3

          [,1]       [,2]        [,3]
[1,] 0.9606359  1.3313254 -0.60785734
[2,] 0.7255531 -0.1389708 -0.02877733

, , 4

          [,1]       [,2]       [,3]
[1,] 0.0279858  0.9007448  0.1251860
[2,] 0.5111250 -0.4223850 -0.6083399

> apply(x,c(1,2),mean) #以第1及第2维为基础，沿第3方向压成平面
           [,1]       [,2]       [,3]
[1,] 0.01790675 0.01413214 -0.2336942
[2,] 0.17291831 0.23199563 -0.3625262

> apply(x,c(1,3),mean)
           [,1]       [,2]     [,3]       [,4]
[1,] -0.4548941 -0.7266532 0.561368  0.3513056
[2,]  0.5149192 -0.4711373 0.185935 -0.1732000

> apply(x,c(2,3),mean)
           [,1]        [,2]       [,3]       [,4]
[1,] -0.6501471 -0.08085267  0.8430945  0.2695554
[2,]  0.7513434 -1.09444509  0.5961773  0.2391799
[3,] -0.0111585 -0.62138791 -0.3183173 -0.2415770