根据来自另一个数据帧的特定行的总和来改变一列

Question

我在这里寻求帮助。我有两个数据框，df1 和 df2。我想根据 df2 中特定行的总和向 df1 添加一个额外的列。

Df1 包含站名。 Df2 包含位置、年份和以度为单位的观测值。我想要每个站的度数总和。这些学位应该是每年特定地点的总和。 可以将其想象为“每个站点都应根据给定的位置，每年获得其度数之和”。我希望只编码站名和位置，desired_output 中的年份应该包括 df2 中给出的所有年份。

失败的示例和所需的输出。我更喜欢在 tidyverse 环境中工作。

一切顺利

df1 <- data.frame(station = c("station_A", "station_B"))

df2 <- data.frame(location= c("south", "north", "north", "east", "west"), year = c(2000, 2000, 2001, 2001, 2001), degrees = c(5,3,9,5,2))

degrees_for_each_station <-
  df1%>% 
  mutate (degrees = case_when(
    station == "station_A" ~ if_else(df2$location %in% c("north","south"),
                                            sum(df2$degrees),
                                            NA),
    station == "station_B" ~ if_else(df2$location %in% c("north","east", "west"),
                                            sum(df2$degrees),
                                            NA)))

desired_output <- data.frame(station = c("station_A", "station_A","station_B", "station_B"),
                             year = c(2000, 2001, 2000, 2001),
                             degrees = c(8,9,3,16))```

Answer 1

一种方法是：

library(tidyverse)

df1 %>%
  left_join(
    df2 %>%
      mutate(
        location = case_when(
          location == 'south' ~ 'station_A',
          location %in% c('east', 'west') ~ 'station_B',
          location == 'north' ~ 'station_A, station_B'
          )
      ) %>%
      separate_rows(location, sep = ', ') %>%
      group_by(location, year) %>%
      summarise(degrees = sum(degrees)),
    by = c('station' = 'location')
  )

输出：

    station year degrees
1 station_A 2000       8
2 station_A 2001       9
3 station_B 2000       3
4 station_B 2001      16