如何在 R 中查找多维数组的模式

Question

我有一个多维数组

 ,,1
    [,1] [,2]        
 [1,] "0"  "a"    
 [2,] "1"  "a"    
 [3,] "1"  "b"
 [2,] "1"  "c"    
 [3,] "1"  "c"
 ,,2
    [,1] [,2]        
 [1,] "0"  "a"    
 [2,] "1"  "a"    
 [3,] "1"  "a"
 [2,] "0"  "b"    
 [3,] "0"  "c"

我只想找到模式，如果每个组数据的第一列为“1”。 对于这个多维数组，我想要 output 可能像这样

 ,,1
 c
 ,,2
 a

或者

[1] c
[2] a    

Answer 1

假设您的数组如下所示：

set.seed(1)
vec <- sample(c(0:3, letters[1:3]), 5 * 2 * 3, replace = TRUE)
dim(vec) <- c(5, 2, 3)
vec
#> , , 1
#> 
#>      [,1] [,2]
#> [1,] "0"  "a" 
#> [2,] "3"  "c" 
#> [3,] "c"  "2" 
#> [4,] "0"  "b" 
#> [5,] "1"  "1" 
#> 
#> , , 2
#> 
#>      [,1] [,2]
#> [1,] "2"  "1" 
#> [2,] "2"  "b" 
#> [3,] "0"  "b" 
#> [4,] "a"  "1" 
#> [5,] "a"  "c" 
#> 
#> , , 3
#> 
#>      [,1] [,2]
#> [1,] "0"  "0" 
#> [2,] "c"  "b" 
#> [3,] "a"  "a" 
#> [4,] "a"  "a" 
#> [5,] "0"  "1"

然后，您可以通过执行以下操作在数组的每个切片中找到最常见的条目：

apply(vec, 3, function(x) names(rev(sort(table(x))))[1])
#> [1] "c" "b" "a"

^{由代表 package (v2.0.1) 于 2022 年 2 月 3 日创建}