CODEWITHSUNDEEP
r
Amo
Amo

Reputation: 15

How to count the number of times a string appears?

I have a dataset like this 

PT <- c(1,1,1,1,2,2,2,3,3,3,3,3)
visit <- c("w1","w2","w3","w4","w1","w2","w3","w1","w2","w3","w4","w5")
dose_level <- c("250", "250", "200", "200", "250", "250", "250", "200","200","150", "150", "100")
test <- data.frame(PT, visit, dose_level)

What I want to do is to calculate the number of times dose-level reduction by each patient. The result would be like:

enter image description here

Thank you in advance. I appreciate all your help.

Upvotes: 0

Views: 49

Answers (4)

Prahlad
Prahlad

Reputation: 138

test <- data.table(test)
test[,dose_level:=as.numeric(as.character(dose_level))]
test[,lag:=shift(dose_level,1),by="PT"]
#calculate difference between actual and lag
test[,diff:=dose_level-lag]
#flag to check if the diff is +ve or -ve 
test[,check:=ifelse(diff<0,1,0)]
test_summarise <- test[,.(sum(check,na.rm=T)),by="PT"]
test_summarise
PT V1
1  1
2  0
3  2

Upvotes: 0

hello_friend
hello_friend

Reputation: 5788

Base R one (obfuscated) liner:

data.frame(PT = unique(PT),
  times = sapply(split(test, test$PT), function(x){sum(c(FALSE, abs(diff(as.numeric(x$dose_level))) > 0))}))

Upvotes: 0

Rui Barradas
Rui Barradas

Reputation: 76402

In base R this can be done with aggregate.

aggregate(dose_level ~ PT, test, function(x){
  y <- as.integer(as.character(x))
  sum(diff(y) < 0)
})
#  PT dose_level
#1  1          1
#2  2          0
#3  3          2

Upvotes: 3

Frank Zhang
Frank Zhang

Reputation: 1688

I guess patient 3 should be 2?

PT <- c(1,1,1,1,2,2,2,3,3,3,3,3)
visit <- c("w1","w2","w3","w4","w1","w2","w3","w1","w2","w3","w4","w5")
dose_level <- c("250", "250", "200", "200", "250", "250", "250", "200","200","150", "150", "100")
test <- data.frame(PT,visit, dose_level)

library(data.table)
setDT(test)[,.(times=sum(dose_level<shift(dose_level),na.rm = TRUE)),by=.(PT)]
#>    PT times
#> 1:  1     1
#> 2:  2     0
#> 3:  3     2

Upvotes: 2

Related Questions