Any seq Experts here?
I want to count the number of values per id. Counting from 1 to n in the correct order works fine. But duplicated values should be labeld with same number.
Any parameter in for seq that i am missing?
Repr. Ex.: Where "count_n" is the value I actually create, and "need" is the desired output.
Thank you in advance. Cheers
Date <- as.Date(c('2006-08-30','2006-08-30','2006-08-23', '2006-09-06',
'2006-09-13', '2006-09-20'))
ID <- c("x1","x1","x1","X2","X3","x1")
need<- c(2,2,1,1,1,3)
df<-data.frame(ID,Date,need)
df<- df%>% arrange(Date)
df$count_n <- ave(as.numeric(df$Date),df$ID,FUN = seq_along)
ID Date need count_n
1 x1 2006-08-23 1 1
2 x1 2006-08-30 2 2
3 x1 2006-08-30 2 3
4 X2 2006-09-06 1 1
5 X3 2006-09-13 1 1
6 x1 2006-09-20 3 4
We can convert Date to factor for each ID which will give you unique value for each Date
ave(as.integer(df$Date),df$ID,FUN = factor)
#[1] 1 2 2 1 1 3
We can also use dense_rank with dplyr
library(dplyr)
df %>%
group_by(ID) %>%
mutate(count_n = dense_rank(Date))
# ID Date need count_n
# <fct> <date> <dbl> <int>
#1 x1 2006-08-23 1 1
#2 x1 2006-08-30 2 2
#3 x1 2006-08-30 2 2
#4 X2 2006-09-06 1 1
#5 X3 2006-09-13 1 1
#6 x1 2006-09-20 3 3