Aggregate date time to summarize time spent at certain conditions

问题

I'm stumped with how I should proceed. I have some dummy data below:

 Date <- as.POSIXct(c('2018-03-20 11:52:25', '2018-03-22 12:01:44', '2018-03-20 12:05:25', '2018-03-20 12:10:40', '2018-03-20 12:12:51 ')) 
Sites<-c(4, 4, 4, 6, 7)
Individual<-c("A", "A", "A", "B", "B")

data.frame(Individual, Date, Sites)

Producing this data frame:

  Individual                Date Sites
          A 2018-03-20 11:52:25     4
          A 2018-03-20 12:01:44     4
          A 2018-03-20 12:05:25     4
          B 2018-03-20 12:10:40     6
          B 2018-03-20 12:12:51     7

For each individual, I would like to determine the time spent at each site. Is there a function that will summarize time spent according to the two conditions? Any help or input is much appreciated.

回答1:

How about this:

library(tidyverse)

Date <- as.POSIXct(c('2018-03-20 11:52:25', '2018-03-22 12:01:44', '2018-03-20 12:05:25', '2018-03-20 12:10:40', '2018-03-20 12:12:51 ')) 
Sites <- c(4, 4, 4, 6, 7)
Individual <- c("A", "A", "A", "B", "B")

df <- data.frame(Individual, Date, Sites)

df %>%
  group_by(Individual, Sites) %>%
  summarise(time_spent = max(Date) - min(Date))
#> # A tibble: 3 x 3
#> # Groups:   Individual [2]
#>   Individual Sites time_spent  
#>   <fct>      <dbl> <time>      
#> 1 A              4 2.00647 days
#> 2 B              6 0.00000 days
#> 3 B              7 0.00000 days

^{Created on 2019-03-20 by the reprex package (v0.2.1)}

回答2:

I'd do...

library(data.table)
setDT(DF)

spellDT = DF[, .(StartDate = first(Date)), by=.(Individual, Site = Sites, g = rleid(Sites))]
spellDT[, duration := shift(StartDate, type="lead") - StartDate, by=Individual][]

   Individual Site g           StartDate      duration
1:          A    4 1 2018-03-20 11:52:25       NA mins
2:          B    6 2 2018-03-20 12:10:40 2.183333 mins
3:          B    7 3 2018-03-20 12:12:51       NA mins

or similarly in dplyr verbs:

library(dplyr)

DF %>% distinct(Individual, g = data.table::rleid(Sites), .keep_all = TRUE) %>% 
  rename(StartDate = Date, Site = Sites) %>%
  group_by(Individual) %>% mutate(duration = lead(StartDate) - StartDate)

# A tibble: 3 x 5
# Groups:   Individual [2]
  Individual StartDate            Site     g duration     
  <fct>      <dttm>              <dbl> <int> <time>       
1 A          2018-03-20 11:52:25     4     1       NA mins
2 B          2018-03-20 12:10:40     6     2 2.183333 mins
3 B          2018-03-20 12:12:51     7     3       NA mins

来源：https://stackoverflow.com/questions/55268381/aggregate-date-time-to-summarize-time-spent-at-certain-conditions

标签

datetime

summary