r - 根据 R 数据帧的每一列中的值所在的范围对它们进行计数

Question

我在 R 中有一个 1246 x 60,660 数据帧。数据片段:

                     gene1             gene2              gene3
sample1          1615.7529         41.932474           697.9728
sample2           663.2001          8.602831          1198.1398
sample3          2406.1532         12.622443          1033.4625
sample4           836.3808         60.144235           259.3720
sample5          1217.8192         22.775497           695.9924
sample6           865.0344         15.350298           683.5397
sample7           935.3658         20.380676           540.6242
sample8           667.3883         56.939874          1056.6981

对于每个基因，我希望将每个样本的值放入以下组之一:

无 = 0

超低 = 1-4

低 = 5 - 100

中 = 101 - 1000

最高 = 1000 及以上

最终产品将是另一个矩阵/数据框，如下所示:

                  gene1      gene2      gene3
none                  0          0          0 
ultra low             0          0          0
low                   0          8          0
medium                5          0          5
high                  3          0          3

我怎样才能做到这一点？经过一番搜索后，我想我可能最终会使用 count 或者 aggregate？但我不确定如何将其应用到每一列。我见过的大多数示例都只计算一列。

Answer 1

使用cut可能会更容易 - 即通过使用循环来指定每列中的breaks和相应的labels lapply 并使用 table

获取频率计数

sapply(df1, \(x) table(cut(x, breaks = c(0, 1, 5, 101, 1001, Inf), 
   labels = c("none", "ultra low", "low", "medium", "high"))))

-输出

           gene1 gene2 gene3
none          0     0     0
ultra low     0     0     0
low           0     8     0
medium        5     0     5
high          3     0     3

或者正如@ZheyuanLi提到的，​​制表可能会更快

lbls <- c("none", "ultra low", "low", "medium", "high")
out <- sapply(df1, \(x) tabulate(cut(x, breaks = c(0, 1, 5, 101, 1001, Inf), 
   labels = lbls), nbins = length(lbls)))
row.names(out) <- lbls

-输出

> out
          gene1 gene2 gene3
none          0     0     0
ultra low     0     0     0
low           0     8     0
medium        5     0     5
high          3     0     3

数据

df1 <- structure(list(gene1 = c(1615.7529, 663.2001, 2406.1532, 836.3808, 
1217.8192, 865.0344, 935.3658, 667.3883), gene2 = c(41.932474, 
8.602831, 12.622443, 60.144235, 22.775497, 15.350298, 20.380676, 
56.939874), gene3 = c(697.9728, 1198.1398, 1033.4625, 259.372, 
695.9924, 683.5397, 540.6242, 1056.6981)), class = "data.frame", 
row.names = c("sample1", 
"sample2", "sample3", "sample4", "sample5", "sample6", "sample7", 
"sample8"))