CODEWITHSUNDEEP
r
DN1
DN1

Reputation: 218

How to conditionally label row sequentially in R?

I am looking to label samples depending on whether their number is 500 more or less than the sample before it. I've seen examples of conditional labelling but I can't find one which fits what I need.

For example my data looks like this: 

column a
200
230
510
1200
1800
1700
2400

I am looking to label each sample depending on if they are close to each other by a maximum of 500. So the output would be:

column a     column b
200            region1
230            region1
510            region1 
1200           region2  #new region starts as there is more than 500 difference than 510 (690)
1400           region2
1700           region2
2400           region3 #new region starts as there is 700 difference from 1700

I've seen examples of conditional labelling, but for all of them there are a set number of labels (for example just binary labels) and I need the label number (region number) to increase with each new region. How can I do this? I have tried adapting other examples but I have made little process in both setting the if more than 500 new label condition and have sequential labelling.

Upvotes: 3

Views: 146

Answers (1)

GKi
GKi

Reputation: 39707

You can use diff and cumsum

x$b  <-  paste0("region", c(1, 1+cumsum(diff(x$a) > 500)))
x
#     a       b
#1  200 region1
#2  230 region1
#3  510 region1
#4 1200 region2
#5 1800 region3
#6 1700 region3
#7 2400 region4

Data

x  <- data.frame(a=c(200,230,510,1200,1800,1700,2400))

Upvotes: 5

Related Questions