r - 如何使用R中的for循环将列保存到单个文件中

Question

我有一个包含 143 列和 147153 行的数据框，如下所示:(我的数据的子集)

    otu1    otu2         1        2        3
1: OTU_1  OTU_10  3.807355 5.403722 3.972693
2: OTU_1 OTU_100 -1.618910 0.000000 0.000000
3: OTU_1 OTU_101  0.000000 0.000000 0.000000

现在我需要 1 和 2 列以及随后的列在单独的文件中。
愿望输出将是:

文件 1:

      otu1     otu2         1
 1:   OTU_1   OTU_10  3.807355
 2:   OTU_1  OTU_100 -1.618910
 3:   OTU_1  OTU_101  0.000000
 4:   OTU_1 OTU_1019  0.000000
 5:   OTU_1  OTU_102  0.000000

文件 2 等等....

      otu1     otu2        2
 1:   OTU_1   OTU_10 5.403722
 2:   OTU_1  OTU_100 0.000000
 3:   OTU_1  OTU_101 0.000000
 4:   OTU_1 OTU_1019 9.211077
 5:   OTU_1  OTU_102 0.000000

所以，我尝试了以下代码:

for ( i in 1:141){
  patient$i=log_trans2[,c(1:2,(i+2))]
  patient$i=patient$i[patient$i!=0]
  patient$i=merge(patient$i, OTUNames, by.x="otu2", by.y="OTUId")
  patient$i=merge(patient$i, OTUNames, by.x="otu1", by.y="OTUId")
  write.table(patient1$i, file=sprintf("patient_grahs/patient.%s.tab",i), sep = '\t',quote = FALSE)
}

但我收到如下错误:

Error in $<-.data.frame(*tmp*, "i", value = c(1, 2, 4)) : replacement has 3 rows, data has 0

我哪里出错了？

Answer 1

我们可以使用 lapply与 fwrite .根据显示的数据，似乎是data.table

library(data.table)
setDT(log_trans2)
lapply(names(log_trans2)[3:ncol(log_trans2)], function(nm) {
     d1 <- log_trans2[, c(names(log_trans2)[1:2], nm), with = FALSE]
     d1 <- merge(d1, OTUNames, by.x="otu2", by.y="OTUId")
     d1 <- merge(d1, OTUNames, by.x="otu1", by.y="OTUId")
   fwrite(d1,  file = paste0("patients_", nm, ".txt"))})

fwrite 的输入将是

lapply(names(log_trans2)[3:ncol(log_trans2)], function(nm) 
     log_trans2[, c(names(log_trans2)[1:2], nm), with = FALSE])
#[[1]]
#    otu1    otu2         1
#1: OTU_1  OTU_10  3.807355
#2: OTU_1 OTU_100 -1.618910
#3: OTU_1 OTU_101  0.000000

#[[2]]
#    otu1    otu2        2
#1: OTU_1  OTU_10 5.403722
#2: OTU_1 OTU_100 0.000000
#3: OTU_1 OTU_101 0.000000

#[[3]]
#    otu1    otu2        3
#1: OTU_1  OTU_10 3.972693
#2: OTU_1 OTU_100 0.000000
#3: OTU_1 OTU_101 0.000000

数据

log_trans2 <- structure(list(otu1 = c("OTU_1", "OTU_1", "OTU_1"), otu2 = c("OTU_10", 
"OTU_100", "OTU_101"), `1` = c(3.807355, -1.61891, 0), `2` = c(5.403722, 
0, 0), `3` = c(3.972693, 0, 0)), class = "data.frame", row.names = c("1:", 
"2:", "3:"))