如何在R中找到不同的时间间隔

Question

I have a vector of dates that looks like this:

"2017-01-05 08:40:00 COT" "2017-01-05 08:50:00 COT" "2017-01-05 09:00:00 COT" "2017-01-05 09:10:00 COT" "2017-01-05 10:30:00 COT" "2017-01-05 10:40:00 COT" "2017-01-05 10:50:00 COT" "2017-01-05 11:00:00 COT" "2017-01-05 11:10:00 COT" "2017-01-05 11:20:00 COT" "2017-01-05 11:30:00 COT" "2017-01-05 11:40:00 COT""2017-01-05 11:50:00 COT" "2017-01-05 12:00:00 COT" "2017-01-05 12:10:00 COT" "2017-01-05 12:20:00 COT" "2017-01-05 12:30:00 COT" "2017-01-05 12:40:00 COT" "2017-01-05 12:50:00 COT" "2017-01-05 13:00:00 COT" "2017-01-05 13:10:00 COT" "2017-01-05 13:20:00 COT" "2017-01-05 13:30:00 COT" "2017-01-05 13:40:00 COT""2017-01-05 13:50:00 COT" "2017-01-05 14:00:00 COT" "2017-01-05 14:10:00 COT" "2017-01-05 14:20:00 COT" "2017-01-05 14:30:00 COT" "2017-01-05 14:40:00 COT" "2017-01-05 14:50:00 COT" "2017-01-05 15:00:00 COT"

and are declare as POSIXct. I need to find the start and end time of the different time intervals in this vector. For this dates there are 2 time intervals: From "2017-01-05 08:40:00 COT" to "2017-01-05 09:10:00 COT" and from "2017-01-05 10:30:00 COT" to "2017-01-05 15:00:00 COT" .

I have tried the command range(data) but i get the starting date and the end date of the whole vector and not the different intervals. Hope someone can help

Answer 1

假设您将所有日期都放在称为“日期”的向量中，则可以执行以下操作

Ranges <- cut(dates, breaks = "2 hours")

Answer 2

If you substitute your text in this input strategy you get a vector of datetimes:

 dat <- as.POSIXct( scan(text=' ... ',quote="\"",what="") )
 # Don't use spaces flanking the ellipsis

Then you could use findInterval to classify the dates and select one with the interval designation you desire, 1 and 3 in your case.

 Breaks <- scan(text='"2017-01-05 08:40:00 COT","2017-01-05 09:10:00 COT","2017-01-05 10:30:00 COT" , "2017-01-05 15:00:00 COT"', sep=",", what="")
# Read 4 items
findInterval(dat,  as.POSIXct(Breaks ) )
# [1] 1 1 1 2 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 4

Answer 3

Using some differences between the times to find the break points in your vector, which I've called tx (code to create below): As you can see, the results split into 8:40 to 9:10 and then 10:30 to the end.

diff(tx)
#Time differences in mins
#[1] 10 10 10 80 10 10 10

cumsum(c(FALSE, diff(tx)!=10))
#[1] 0 0 0 0 1 1 1 1

split(tx, cumsum(c(FALSE, diff(tx)!=10)))
#$`0`
#[1] "2017-01-05 08:40:00 COT" "2017-01-05 08:50:00 COT" "2017-01-05 09:00:00 COT" "2017-01-05 09:10:00 COT"
#
#$`1`
#[1] "2017-01-05 10:30:00 COT" "2017-01-05 10:40:00 COT" "2017-01-05 10:50:00 COT" "2017-01-05 11:00:00 COT"

Where tx was:

tx <- structure(c(1483623600, 1483624200, 1483624800, 1483625400, 1483630200, 
1483630800, 1483631400, 1483632000), class = c("POSIXct", "POSIXt"
), tzone = "America/Bogota")