CODEWITHSUNDEEP
rdplyr
frank kruger
frank kruger

Reputation: 25

Counting groups in R(dplyr package) based on 2 statements

I would like to know if it is possible using dplyr to count the number of "visits". A visit is defined by every time the elapsed time is more than 30 or when the species change.

I tried grouping them by species 

visit <- rawdata %>%
  group_by(Species)%>%
  mutate(VisitNo = cumsum(Elapsed >= 30)

But this makes that every species starts again at 0

Species    Elapsed VisitNo
aardvark     5 
aardvark     10
aardvark     2 
aardvark     30 
aardvark     4 
aardvark     30 
aardvark     10
Jackal       5
Jackal       30
Impala       5
Impala       30`

expected output

Species    Elapsed VisitNo
aardvark     5          1
aardvark     10         1
aardvark     2          1
aardvark     30         2
aardvark     4          2 
aardvark     30         3
aardvark     10         3
Jackal       5          4
Jackal       30         5
Impala       5          5
Impala       30         6

Thanks for the help.

Upvotes: 1

Views: 29

Answers (2)

Ronak Shah
Ronak Shah

Reputation: 388797

Another option : 

library(dplyr)

df %>% 
   mutate(VisitNo = cumsum(Species != lag(Species, default = last(Species)) | 
                           Elapsed >= 30))

#    Species Elapsed VisitNo
#1  aardvark       5       1
#2  aardvark      10       1
#3  aardvark       2       1
#4  aardvark      30       2
#5  aardvark       4       2
#6  aardvark      30       3
#7  aardvark      10       3
#8    Jackal       5       4
#9    Jackal      30       5
#10   Impala       5       6
#11   Impala      30       7

Similarly with data.table

library(data.table)
setDT(df)[, VisitNo := cumsum(Species != shift(Species, fill = last(Species)) | 
                               Elapsed >= 30)]

Upvotes: 1

tmfmnk
tmfmnk

Reputation: 39858

You can do:

df %>%
 mutate(VisitNo = cumsum(!duplicated(Species) | Elapsed >= 30))

    Species Elapsed VisitNo
1  aardvark       5       1
2  aardvark      10       1
3  aardvark       2       1
4  aardvark      30       2
5  aardvark       4       2
6  aardvark      30       3
7  aardvark      10       3
8    Jackal       5       4
9    Jackal      30       5
10   Impala       5       6
11   Impala      30       7

Upvotes: 2

Related Questions