CODEWITHSUNDEEP
rlubridate
Mark Heckmann
Mark Heckmann

Reputation: 11431

calculate duration of time interval while removing certain time spans

Assume we have an interval spanning several days (interval "A" in Figure below).

library(lubridate)
int <- interval("2018-01-01 22:00:00", "2018-01-04 10:00:00")

In hours, I get

as.period(int, unit = "hours")
"60H 0M 0S"

Now, I want to subtract all non-working-hours, here 16:00-08:00 (greyed out) in that interval, i.e. only keep the blue parts (08:00-16:00) and, again, calculate the remaining hours (see "B" in Figure below), which would be 8 + 8 + 2 = 18 hours.

enter image description here

One approach would be to create a list of intervals I want to keep which span the entire interval and then calculate intersections. (The code below could, of course, be setup programmatically using floor/ceiling/seq functions etc.)

int_keep <- list(
  interval("2018-01-01 08:00:00", "2018-01-01 16:00:00"),
  interval("2018-01-02 08:00:00", "2018-01-02 16:00:00"),
  interval("2018-01-03 08:00:00", "2018-01-03 16:00:00"),
  interval("2018-01-04 08:00:00", "2018-01-04 16:00:00"),
  interval("2018-01-05 08:00:00", "2018-01-05 16:00:00")
)

l <- lapply(int_keep, function(x) intersect(x, int))
mns <- sapply(l, as.numeric)    # returns seconds
sum(mns, na.rm = T) / 60 / 60   # sum of intersections in hours
[1] 18

While this works, it appears utterly clumsy to me. What would be a less tedious way to do this?

Upvotes: 1

Views: 141

Answers (1)

Shirin Yavari
Shirin Yavari

Reputation: 682

df <- data.frame(DateTime=seq.POSIXt(as.POSIXct("2018-01-01 22:00:00"), as.POSIXct("2018-01-04 10:00:00"), by = "1 hour"))
head(df)
#DateTime
#1 2018-01-01 22:00:00
#2 2018-01-01 23:00:00
#3 2018-01-02 00:00:00
#4 2018-01-02 01:00:00
#5 2018-01-02 02:00:00
#6 2018-01-02 03:00:00

#you want the hours worked between A and B
A <-format(strptime("8:00:00", "%H:%M:%S"),"%H:%M:%S")
B <-format(strptime("16:00:00", "%H:%M:%S"),"%H:%M:%S")
#a simple ifelse statement to assign a value of 1 to column "value" if the time is between 8 and 16 or a 0 if it's not: 
df$value<-ifelse((format(df[1],"%H:%M:%S")>A & format(df[1],"%H:%M:%S")<=B),1,0)
 tail(df)
 #DateTime DateTime
 #56 2018-01-04 05:00:00        0
 #57 2018-01-04 06:00:00        0
 #58 2018-01-04 07:00:00        0
 #59 2018-01-04 08:00:00        0
 #60 2018-01-04 09:00:00        1
 #61 2018-01-04 10:00:00        1
#now taking the column sum of the value column will give you the total hours worked: 
TotalHoursWorked<-colSums(df$value)
TotalHoursWorked
#DateTime 
 # 18

Upvotes: 1

Related Questions