在循环内粘贴指定的 3 列

Question

df<-data.frame(expand.grid(c("a","b","c"),c("p","q","r"),c("x","y","z"),c("l","m","n")));

我有这个有 4 列的表，我想将结果放在一个向量中，这样它应该是

paste0(df$Var1,df$Var2,df$Var4)

这里提出的问题只是为了演示目的，所以我希望它非常动态和灵活。

我正在寻找任何可以按指定连接 2、3 或 4 列的代码。

如果我们说

i<-1;
j<-2;
k<-4;
paste0(df[,i],df[,j],df[,k])

现在想象一下，如果 df 中有很多列，则执行相同的操作。还有列数以及哪些列应基于输入。

如果这是使用 data.table 包完成的，我将不胜感激。

Answer 1

我想你需要do.call。

df <- expand.grid(c("a","b","c"),c("p","q","r"),c("x","y","z"))
do.call(paste0, df)
#  [1] "apx" "bpx" "cpx" "aqx" "bqx" "cqx" "arx" "brx" "crx" "apy" "bpy" "cpy" "aqy" "bqy" "cqy" "ary" "bry" "cry" "apz" "bpz" "cpz" "aqz" "bqz" "cqz" "arz"
# [26] "brz" "crz"

既然你提到了data.table，那么让我构建正确的对象：

library(data.table)
# option 1
df <- data.table(expand.grid(c("a","b","c"),c("p","q","r"),c("x","y","z")))
# option 2
df <- expand.grid(c("a","b","c"),c("p","q","r"),c("x","y","z"))
setDT(df)

# then
df[, do.call(paste0, .SD)]
#  [1] "apx" "bpx" "cpx" "aqx" "bqx" "cqx" "arx" "brx" "crx" "apy" "bpy" "cpy" "aqy" "bqy" "cqy" "ary" "bry" "cry" "apz" "bpz" "cpz" "aqz" "bqz" "cqz" "arz"
# [26] "brz" "crz"

使用data.table's，.SD您可以指定某些列，例如df[, do.call(paste0, .SD), .SDcols = c(i, j, k)]。

您需要选择的列只是一个列子集 ala df[,c(1,2,4)]。

df <- expand.grid(c("a","b","c"),c("p","q","r"),c("x","y","z"),c("l","m","n"))
i <- 1
j <- 2
k <- 4
do.call(paste0, df[,c(i, j, k)])
#  [1] "apl" "bpl" "cpl" "aql" "bql" "cql" "arl" "brl" "crl" "apl" "bpl" "cpl" "aql" "bql" "cql" "arl" "brl" "crl" "apl" "bpl" "cpl" "aql" "bql" "cql" "arl"
# [26] "brl" "crl" "apm" "bpm" "cpm" "aqm" "bqm" "cqm" "arm" "brm" "crm" "apm" "bpm" "cpm" "aqm" "bqm" "cqm" "arm" "brm" "crm" "apm" "bpm" "cpm" "aqm" "bqm"
# [51] "cqm" "arm" "brm" "crm" "apn" "bpn" "cpn" "aqn" "bqn" "cqn" "arn" "brn" "crn" "apn" "bpn" "cpn" "aqn" "bqn" "cqn" "arn" "brn" "crn" "apn" "bpn" "cpn"
# [76] "aqn" "bqn" "cqn" "arn" "brn" "crn"

正如 GregorThomas 在评论中所说，其data.table- 形式是：

as.data.table(df)[, do.call(paste0, .SD), .SDcols = c(i, j, k)]

Answer 2

另一种选择是使用Reduce()：

df <-
  data.frame(expand.grid(c("a", "b", "c"), c("p", "q", "r"), c("x", "y", "z")))

df_pasted <- Reduce(paste0, df)
df_pasted
#>  [1] "apx" "bpx" "cpx" "aqx" "bqx" "cqx" "arx" "brx" "crx" "apy" "bpy" "cpy"
#> [13] "aqy" "bqy" "cqy" "ary" "bry" "cry" "apz" "bpz" "cpz" "aqz" "bqz" "cqz"
#> [25] "arz" "brz" "crz"

^{创建于 2023-08-28，使用reprex v2.0.2}

编辑：正如 r2evans 所指出的，您可以只指定要保留的列。如果您有data.table可以使用：

Reduce(paste0, df[, .(i, j, k), with = FALSE])