根据 R 中另一个数据帧的时间范围更改 id？

Question

这是我的数据框：

sensors_data <- structure(list(timestamp = c(164424852623558, 164424852623558, 
164424855959288, 164424855959288, 164424857931288, 164424857931288, 
164424859915288, 164424859915288, 164424861778558, 164424861903788
), acc_x = c(5.4230323, 5.4230323, 5.4086666, 5.4086666, 5.4086666, 
5.4086666, 5.3895125, 5.3895125, 5.3895125, 5.3895125), acc_y = c(7.9681463, 
7.9681463, 7.9777236, 7.9777236, 7.9777236, 7.9777236, 7.987301, 
7.987301, 7.987301, 7.987301), acc_z = c(2.6097596, 2.6097596, 
2.6337023, 2.6337023, 2.6337023, 2.6337023, 2.6528566, 2.6528566, 
2.6528566, 2.6528566), grav_x = c(-1, -1, -1, -1, -1, -1, -1, 
-1, 5.82316, 5.82316), grav_y = c(-1, -1, -1, -1, -1, -1, -1, 
-1, 7.7164946, 7.7164946), grav_z = c(-1, -1, -1, -1, -1, -1, 
-1, -1, 1.6482342, 1.6482342), gyro_x = c(-1, -0.3230286, -0.3230286, 
-0.3230286, -0.3230286, -0.3230286, -0.3230286, -0.3230286, -0.3230286, 
-0.13915816), gyro_y = c(-1, 0.77723867, 0.77723867, 0.77723867, 
0.77723867, 0.77723867, 0.77723867, 0.77723867, 0.77723867, 1.0466303
), gyro_z = c(-1, -0.365287, -0.365287, -0.365287, -0.365287, 
-0.365287, -0.365287, -0.365287, -0.365287, -0.57420295), press_id = c(-1, 
-1, -1, -1, -1, -1, -1, -1, -1, -1)), .Names = c("timestamp", 
"acc_x", "acc_y", "acc_z", "grav_x", "grav_y", "grav_z", "gyro_x", 
"gyro_y", "gyro_z", "press_id"), row.names = c(NA, -10L), class = c("tbl_df", 
"tbl", "data.frame"))

我想根据其他数据框将 press_id 变异/更改为特定值：

pin_press_time_range <- structure(list(press_id = 1:4, start_time = c(164429106370978, 
164429411618824, 164429837271939, 164430399454284), end_time = c(164429182443824, 
164429512525747, 164429903243169, 164430465927554)), class = c("tbl_df", 
"tbl", "data.frame"), .Names = c("press_id", "start_time", "end_time"
), row.names = c(NA, -4L))

这意味着对于 press_id 1、2、... 的开始时间和结束时间之间的所有时间戳，根据 press_id 进行标记。

我正在尝试使用 mutate_if 或 mutate_at 执行此操作，但没有运气。

这就是我手动执行此操作的方式，但我需要它自动执行：

sensors_data %>% filter(timestamp >= pin_press_time_range[1,]$start_time & 
           timestamp <= pin_press_time_range[1,]$end_time) %>%
  mutate(press_id = pin_press_time_range[1,]$press_id)

请指教。

Answer 1

使用data.table::foverlaps

library(data.table)
setDT(sensors_data)[, `:=`(start_time = timestamp, end_time = timestamp)]
setDT(pin_press_time_range)

setkey(pin_press_time_range, start_time, end_time)
dt <- foverlaps(sensors_data, pin_press_time_range)

解释：foverlaps 执行重叠连接，基于两个data.tables 的间隔；由于foverlaps 需要起点和终点，我们选择timestamp 作为sensors_data 的起点和。然后，我们根据start_time 和end_time 两个公共键离开重叠连接sensors_data 和pin_press_time_range。

Answer 2

使用sqldf

library(sqldf)
sqldf("select a.*, b.*
      from sensors_data a
          left join pin_press_time_range b
           on a.timestamp >= b.start_time 
           AND a.timestamp <= b.end_time")

Answer 3

使用模糊连接

library(fuzzyjoin)    

sensors_data %>%
  fuzzy_left_join(pin_press_time_range,
                  by = c("timestamp" = "start_time", "timestamp" = "end_time"), 
                  match_fun = list(`>=`, `<=`))