r - 根据另一个 df 替换选定行中的多个列值

Question

我正在寻找一种迭代方式（例如 for loop、purrr）用另一个数据帧中类似列的值替换数据帧选定行的某些列的值。

# Here is a toy df
df <- dplyr::tibble( 
  "id" = c("id_a", "id_b", "id_c"), 
  "subj_1" = c("subj_1_a", "subj_1_b",  "subj_1_c"), 
  "subj_2" = c("subj_2_a", "subj_2_b","subj_2_c"), 
  "subj_3" =  c("subj_3_a", "subj_3_b","subj_3_c") 
  )

#   id    subj_1   subj_2   subj_3  
#   <chr> <chr>    <chr>    <chr>   
# 1 id_a  subj_1_a subj_2_a subj_3_a
# 2 id_b  subj_1_b subj_2_b subj_3_b
# 3 id_c  subj_1_c subj_2_c subj_3_c

这是一个数据名，其中包含我想在列中替换的值：

df_replace <- dplyr::tibble(
  "id" = "id_a",  
  "subj_1" = "subj_1_a_REP",
  "subj_2" = "subj_2__REP",
  "subj_3" = "subj_2__REP"
)

这是我求的结果（一一得到）：

df2 <- df
df2$subj_1[df2$id == "id_a"] <- df_replace$subj_1
df2$subj_2[df2$id == "id_a"] <- df_replace$subj_2
df2$subj_3[df2$id == "id_a"] <- df_replace$subj_3

#   id    subj_1       subj_2      subj_3     
#   <chr> <chr>        <chr>       <chr>      
# 1 id_a  subj_1_a_REP subj_2__REP subj_2__REP
# 2 id_b  subj_1_b     subj_2_b    subj_3_b   
# 3 id_c  subj_1_c     subj_2_c    subj_3_c

如何获得与 loop 或 purrr 相同的迭代？

我的尝试无效 - 我认为它存在一些索引问题

col_subj_names <- names(df_replace)[-1] # without id

# pick the id 
i <- "id_a"

# for loop (NOT WORKING!)
for (j in seq_along(col_subj_names)) {
  col <-  col_subj_names[j]
    df %>%
       dplyr::filter(id == i) %>%
       dplyr::mutate(., col = df_replace$col)
    df3 <- df
}

我也看不懂警告信息：

Unknown or uninitialized column: col.

Answer 1

这里根本不需要循环。您可以一次性替换整行。

df[df$id %in% df_replace$id,] <- df_replace

df
#> # A tibble: 3 x 4
#>   id    subj_1       subj_2      subj_3     
#>   <chr> <chr>        <chr>       <chr>      
#> 1 id_a  subj_1_a_REP subj_2__REP subj_2__REP
#> 2 id_b  subj_1_b     subj_2_b    subj_3_b   
#> 3 id_c  subj_1_c     subj_2_c    subj_3_c

请注意，即使您有多个 id_a 行，这仍然有效 - 它们都将被适当地替换为单行代码。

Answer 2

尝试 rows_update(df_replace):

library(tidyverse)

df_replace <- tibble(
  "id" = "id_a",  
  "subj_1" = "subj_1_a_REP",
  "subj_2" = "subj_2__REP",
  "subj_3" = "subj_2__REP"
)

tibble( 
  "id" = c("id_a", "id_b", "id_c"), 
  "subj_1" = c("subj_1_a", "subj_1_b",  "subj_1_c"), 
  "subj_2" = c("subj_2_a", "subj_2_b","subj_2_c"), 
  "subj_3" =  c("subj_3_a", "subj_3_b","subj_3_c") 
) |> 
  rows_update(df_replace)
#> Matching, by = "id"
#> # A tibble: 3 × 4
#>   id    subj_1       subj_2      subj_3     
#>   <chr> <chr>        <chr>       <chr>      
#> 1 id_a  subj_1_a_REP subj_2__REP subj_2__REP
#> 2 id_b  subj_1_b     subj_2_b    subj_3_b   
#> 3 id_c  subj_1_c     subj_2_c    subj_3_c

^{由 reprex package (v2.0.1)}

于 2022-05-17 创建

r - 根据另一个 df 替换选定行中的多个列值

r - replacing multiple column values in selected rows based on another df

r

for-loop

purrr