Question

我有一个数据，其中一些列定义组，一些列（下面的示例数据中的a1-a4）只有一列中的值，其余列中有NA。

structure(list(gp = structure(c(1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L), .Label = "gp1", class = "factor"), id = c(1, 1, 2, 2, 2, 2, 3, 3, 3), name = structure(c(1L, 1L, 2L, 2L, 2L, 2L, 3L, 3L, 3L), .Label = c("A", "B", "C"), class = "factor"), a1 = c(0.4, NA, NA, NA, NA, NA, 0.3, NA, NA), a2 = c(NA, NA, NA, 1, NA, NA, NA, NA, NA), a3 = c(NA, 1.2, NA, NA, NA, NA, NA, NA, NA), a4 = c(NA, NA, 1, NA, NA, NA, NA, NA, 1)), .Names = c("gp", "id", "name", "a1", "a2", "a3", "a4"), row.names = c(NA, -9L), class = "data.frame")

据我所知，只有列a1中的一个实际上有价值而且我不需要单独的行，我想将组内的所有值收集到一行。我期待下面的内容。

structure(list(gp = structure(c(1L, 1L, 1L), .Label = "gp1", class = "factor"), id = c(1, 2, 3), name = structure(1:3, .Label = c("A", "B", "C"), class = "factor"), a1 = c(0.4, NA, 0.3), a2 = c(NA, 1, NA), a3 = c(1.2, NA, NA), a4 = c(NA, 1, 1)), .Names = c("gp", "id", "name", "a1", "a2", "a3", "a4"), row.names = c(NA, -3L), class = "data.frame")

我怎样才能实现这一目标？如果解决方案使用tidyverse会很棒。

Answer 1

你可以试试这个

library(tidyverse)
df1 %>% 
 group_by(gp, id, name) %>% 
 summarise_all(sum, na.rm = TRUE) %>% 
 summarise_all(na_if, 0)
# A tibble: 3 x 7
# Groups:   gp [?]
#  gp       id name      a1    a2    a3    a4
#  <fct> <dbl> <fct>  <dbl> <dbl> <dbl> <dbl>
#1 gp1      1. A      0.400   NA   1.20   NA 
#2 gp1      2. B     NA        1. NA       1.
#3 gp1      3. C      0.300   NA  NA       1.

最终输出中不会有NA个但0 s，因此第二次调用summarise_all。我在此假设0到a1列中没有a4。

这是针对初始数据集中有0个的情况的解决方案。

sum_NA <- function(x) {
  if(all(is.na(x))) {
    NA
  } else {
    sum(x, na.rm = TRUE)
  }
}

df2 %>% 
 group_by(gp, id, name) %>% 
 summarise_all(sum_NA)
# A tibble: 3 x 7
# Groups:   gp, id [?]
#  gp       id name      a1    a2    a3    a4
#  <fct> <dbl> <fct>  <dbl> <dbl> <dbl> <dbl>
#1 gp1      1. A      0.      NA   1.20   NA 
#2 gp1      2. B     NA        0. NA       1.
#3 gp1      3. C      0.300   NA  NA       1.

数据

df1 <- structure(list(gp = structure(c(1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L), .Label = "gp1", class = "factor"), id = c(1, 1, 2, 2, 2, 2, 3, 3, 3), name = structure(c(1L, 1L, 2L, 2L, 2L, 2L, 3L, 3L, 3L), .Label = c("A", "B", "C"), class = "factor"), a1 = c(0.4, NA, NA, NA, NA, NA, 0.3, NA, NA), a2 = c(NA, NA, NA, 1, NA, NA, NA, NA, NA), a3 = c(NA, 1.2, NA, NA, NA, NA, NA, NA, NA), a4 = c(NA, NA, 1, NA, NA, NA, NA, NA, 1)), .Names = c("gp", "id", "name", "a1", "a2", "a3", "a4"), row.names = c(NA, -9L), class = "data.frame") df2 <- structure(list(gp = structure(c(1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L), .Label = "gp1", class = "factor"), id = c(1, 1, 2, 2, 2, 2, 3, 3, 3), name = structure(c(1L, 1L, 2L, 2L, 2L, 2L, 3L, 3L, 3L), .Label = c("A", "B", "C"), class = "factor"), a1 = c(0.0, NA, NA, NA, NA, NA, 0.3, NA, NA), a2 = c(NA, NA, NA, 0, NA, NA, NA, NA, NA), a3 = c(NA, 1.2, NA, NA, NA, NA, NA, NA, NA), a4 = c(NA, NA, 1, NA, NA, NA, NA, NA, 1)), .Names = c("gp", "id", "name", "a1", "a2", "a3", "a4"), row.names = c(NA, -9L), class = "data.frame")

Answer 2

由于OP已经提到只有一行具有值，因此在应用dplyr::first后，一个选项可以是group_by。我更倾向于使用summarise_at来灵活地排除几个不需要分析的列。

library(dplyr)

df %>% group_by(gp, id, name) %>%
  summarise_at(vars(starts_with("a")), funs(dplyr::first(sort(.)))) %>%
  as.data.frame()

#    gp id name  a1 a2  a3 a4
# 1 gp1  1    A 0.4 NA 1.2 NA
# 2 gp1  2    B  NA  1  NA  1
# 3 gp1  3    C 0.3 NA  NA  1

Answer 3

dat%>%
  group_by(gp,id,name)%>%
  summarise_all(funs(lift(coalesce)(.)))

# A tibble: 3 x 7
# Groups:   gp, id [?]
  gp       id name      a1    a2    a3    a4
  <fct> <dbl> <fct>  <dbl> <dbl> <dbl> <dbl>
1 gp1      1. A      0.400   NA   1.20   NA 
2 gp1      2. B     NA        1. NA       1.
3 gp1      3. C      0.300   NA  NA       1.

在R中绑定组内的行

3 个答案: