通过比较R中一列中的值来选择行[重复]

Question

我有一个看起来像这样的数据框：

n4= 
    sector turb    dist
    1  sector1  T02  828.66
    2  sector1  T04 1114.58
    3  sector1  T05 1012.22
    4  sector2  T03  992.64
    5  sector2  T05 1012.22
    6  sector2  T06 1158.38
    7  sector3  T03  992.64
    8 sector12  T02  828.66
    9 sector12  T04 1114.58

我想保留具有唯一扇区名称的行，并且措施是在 dist 列中保留具有最小值的行：

 sector turb   dist
1  sector1  T02 828.66
4  sector2  T03 992.64
7  sector3  T03 992.64
8 sector12  T02 828.66

我知道我必须根据部门对它们进行分组：

result = n4 %>%
dplyr::group_by(sector)

但后来使用 select 或 filter 命令并没有像我尝试的那样工作：

result = n4 %>%
    dplyr::group_by(sector)%>%
    dplyr::select(which.min(dist))

知道我该怎么做吗？

Answer 1

您可以使用filter 代替select 并作为slice 的替代品

> n4 %>%
    dplyr::group_by(sector)%>%
    dplyr::filter(dist==min(dist))
# A tibble: 4 x 3
# Groups:   sector [4]
  sector   turb   dist
  <fct>    <fct> <dbl>
1 sector1  T02    829.
2 sector2  T03    993.
3 sector3  T03    993.
4 sector12 T02    829.

如果您更喜欢使用 R 基础，请尝试使用 aggregate

> aggregate(.~sector, data=n4, min)
    sector turb   dist
1  sector1    1 828.66
2 sector12    1 828.66
3  sector2    2 992.64
4  sector3    2 992.64

您可以check this answer 获取更多替代方案来完成此任务。

Answer 2

我们需要slice 而不是select 来对行进行子集化。 select 函数是选择数据集的列。如果“扇区”的顺序应与输入数据中“扇区”的出现顺序相同，则将列更改为factor，并在输入数据顺序中指定levels

n4 %>%       
   dplyr::group_by(sector = factor(sector, levels = unique(sector)))%>%
   dplyr::slice(which.min(dist))
# A tibble: 4 x 3
# Groups:   sector [4]
#  sector   turb   dist
#  <fct>    <chr> <dbl>
#1 sector1  T02    829.
#2 sector2  T03    993.
#3 sector3  T03    993.
#4 sector12 T02    829.

---

或使用base R

n4[with(n4, ave(dist, sector, FUN = min) == dist),]
#     sector turb   dist
#1  sector1  T02 828.66
#4  sector2  T03 992.64
#7  sector3  T03 992.64
#8 sector12  T02 828.66

数据

n4 <- structure(list(sector = c("sector1", "sector1", "sector1", "sector2", 
"sector2", "sector2", "sector3", "sector12", "sector12"), turb = c("T02", 
"T04", "T05", "T03", "T05", "T06", "T03", "T02", "T04"), dist = c(828.66, 
 1114.58, 1012.22, 992.64, 1012.22, 1158.38, 992.64, 828.66, 1114.58
)), class = "data.frame", row.names = c("1", "2", "3", "4", "5", 
"6", "7", "8", "9"))