Question

这是我的数据框：

dput(test)
structure(list(Branch = c("11 Oktomvri", "11 Oktomvri", "11 Oktomvri", 
"11 Oktomvri", "11 Oktomvri", "11 Oktomvri", "11 Oktomvri", "Aerodrom", 
"Aerodrom", "Aerodrom", "Aerodrom", "Aerodrom", "Aerodrom", "Aerodrom", 
"Aerodrom 2", "Aerodrom 2", "Aerodrom 2", "Aerodrom 2", "Aerodrom 2", 
"Aerodrom 2", "Aerodrom 2", "Bitola", "Bitola", "Bitola", "Bitola", 
"Bitola", "Bitola", "Bitola"), period = c("January", "February", 
"March", "April", "May", "June", "July", "January", "February", 
"March", "April", "May", "June", "July", "January", "February", 
"March", "April", "May", "June", "July", "January", "February", 
"March", "April", "May", "June", "July"), value = c(1513, 1511, 
1520, 1524, 1508, 1504, 1517, 1364, 1381, 1400, 1403, 1401, 1406, 
1430, 674, 687, 689, 690, 696, 705, 715, 4400, 4393, 4365, 4342, 
4345, 4373, 4389)), .Names = c("Branch", "period", "value"), row.names = c(NA, 
-28L), class = c("grouped_df", "tbl_df", "tbl", "data.frame"), vars = c("Branch", 
"period"), drop = TRUE, indices = list(3L, 1L, 0L, 6L, 5L, 2L, 
    4L, 10L, 8L, 7L, 13L, 12L, 9L, 11L, 17L, 15L, 14L, 20L, 19L, 
    16L, 18L, 24L, 22L, 21L, 27L, 26L, 23L, 25L), group_sizes = c(1L, 
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L), biggest_group_size = 1L, labels = structure(list(
    Branch = c("11 Oktomvri", "11 Oktomvri", "11 Oktomvri", "11 Oktomvri", 
    "11 Oktomvri", "11 Oktomvri", "11 Oktomvri", "Aerodrom", 
    "Aerodrom", "Aerodrom", "Aerodrom", "Aerodrom", "Aerodrom", 
    "Aerodrom", "Aerodrom 2", "Aerodrom 2", "Aerodrom 2", "Aerodrom 2", 
    "Aerodrom 2", "Aerodrom 2", "Aerodrom 2", "Bitola", "Bitola", 
    "Bitola", "Bitola", "Bitola", "Bitola", "Bitola"), period = c("April", 
    "February", "January", "July", "June", "March", "May", "April", 
    "February", "January", "July", "June", "March", "May", "April", 
    "February", "January", "July", "June", "March", "May", "April", 
    "February", "January", "July", "June", "March", "May")), row.names = c(NA, 
-28L), class = "data.frame", vars = c("Branch", "period"), drop = TRUE, .Names = c("Branch", 
"period")))

我不确定如何根据Branch和Period对数据进行分组，并根据period_n-period_n-1计算值。

输出应为：

city     period   value  diff_n_1
Bitola   March     4365    -28
Bitola   April     2000    13

我的观察：

results <- sample2 %>% 
  group_by(Branch, period) %>%
  arrange(Branch) %>% 
  mutate(lagged_period = lag(value), client_diff = value - lagged_period)

我不确定如何拉下最后一行。

有什么想法吗？

Answer 1

我认为你几乎没有，但你不需要Period中的group_by，因为这确实是我们使用lag的变量。目前，您的上述工作仅产生NAs，因为每个定义的组中只有一个元素（因此没有可用的延迟）。

这应该有效：

library(dplyr)

sample2 %>% 
  group_by(Branch) %>%
  arrange(Branch) %>% 
  mutate(lagged_period = lag(value),
         client_diff = value - lagged_period)

然后，当然，如果您想从计算字段中删除NAs，您可以将上述内容输入：

filter(!is.na(client_diff))

分组数据并计算期间的n-1

1 个答案: