Create a custom ifelse based in proportion limit values

November 25, 2024

I’d like to create some trick using a custom ifelse based in a rule comparing dates. If the value in the next date is 10% higher
comparing with later date then repeat the old value, if not accept the new value. Using artificial values:

# Packages
library(dplyr)
library(lubridate)

# Dataset
area <- c("A","A","A","B","B","B","C","C","C")
dates <- c("05-01-2024","10-01-2024","18-01-2024",
"05-01-2024","10-01-2024","18-01-2024","05-01-2024",
"10-01-2024","18-01-2024")
response <- c(63,68,82,77,71,60,79,75,73)
my.ds <- data.frame(area=area,dates=dates,response=response)

# 
max.prop <- 0.1
my.ds$dates <- lubridate::dmy(my.ds$dates)

# Here I don't have success in using ifelse invoke values at a later date. 
# My output must be:

# my.ds.new
#  area      dates response
#1    A 2024-01-05       63
#2    A 2024-01-10       68
#3    A 2024-01-18       68
#4    B 2024-01-05       77
#5    B 2024-01-10       71
#6    B 2024-01-18       71
#7    C 2024-01-05       79
#8    C 2024-01-10       75
#9    C 2024-01-18       73

Plaese, any help wit it?

>Solution :

my.ds.new <- my.ds %>% mutate(response = ifelse(c(0, abs(diff(response))) > abs(response * .1) & 
                                                  lag(area) == area, 
                                                lag(response), 
                                                response))