在 data.table 中创建子组

Question

假设我有以下简化数据集：

dt <- data.table(id = 1:5, val = c(1, 2, 3, 2, 4))
dt2 <- data.table(id = c(2, 4), val = c(2, 3))

我想替换 dt 中所有值为 2 的值。替换值在 dt2 中给出。这两个表可以通过 id 连接。如果值不等于 2，则最终值应保持不变。如果等于 2，则应变为 paste0(dt$val, ".", dt2$val).

期望输出：

row id val
1:  1   1
2:  2   2.2
3:  3   3
4:  4   2.3
5:  5   4

我试过的（有效但似乎不够优雅）：

merged <- merge(x = dt, y = dt2, by= "id", all.x = TRUE)
merged[!is.na(merged$val.y), ]$val.x <- paste0(
  merged[!is.na(merged$val.y), ]$val.x, ".",
  merged[!is.na(merged$val.y), ]$val.y)
merged[, val.y := NULL]
setnames(x = merged, old = "val.x", new = "val")
merged

问题：如何更优雅地进行转换？

Answer 1

library(data.table)

# example data
dt <- data.table(id = 1:5, val = c(1, 2, 3, 2, 4))
dt2 <- data.table(id = c(2, 4), val = c(2, 3))

如果你的数据集都是正确排序的，你可以像这样使用 base R:

dt$val[dt$id %in% dt2$id] = paste0(dt$val[dt$id %in% dt2$id], ".", dt2$val)

dt

#    id val
# 1:  1   1
# 2:  2 2.2
# 3:  3   3
# 4:  4 2.3
# 5:  5   4

否则你可以使用这个：

dt_merged = merge(dt, dt2, by="id", all.x=T)[, val:=ifelse(is.na(val.y), 
                                                           val.x, 
                                                           paste0(val.x, ".", val.y))]
dt_merged = dt_merged[, c("id","val")]
dt_merged

#    id val
# 1:  1   1
# 2:  2 2.2
# 3:  3   3
# 4:  4 2.3
# 5:  5   4

Answer 2

你正在寻找更新加入

dt[dt2, on=.(id), val := paste0(x.val, ".", i.val)]

输出：

   id val
1:  1   1
2:  2 2.2
3:  3   3
4:  4 2.3
5:  5   4

数据：

#val column needs to be of character type to suppress the warning
dt <- data.table(id = 1:5, val = as.character(c(1, 2, 3, 2, 4)))
dt2 <- data.table(id = c(2, 4), val = c(2, 3))

在 data.table 中创建子组

Create subgroups in data.table

r

melt

data.table