Dynamic mutate case_when according to string matching

66 Views Asked by At

I have a vector which stores all possible merging combinations for a qualitative variable.
The vector has the following structure :

Config = c("4 With 1 + 6 With 2 + 8 With 3","4 With 1 With 2 + 6 With 8")

Which means that the variable with the first value of Config should have 3 modalities : Values 4 With 1 , Values 6 With 2 , Values 8 With 3

I'd like to assign a dynamic value to every element of this merging in a dynamic way, such as :

mtcars %>% mutate(Merging=case_when(carb %in% c(4,1)~"Group 1", carb %in% c(6,2)~"Group 2",carb %in% c(8,3) ~"Group 3"))

My main difficulty here is that the numbers of groups is not the same for each element of config : For one configuration, it could be 3 groups, for another one 4, etc.

I've already tried to work with string match but this possibility only works for 1 group that contains only the word "With" :

mutate(Merging= case_when(
      !!sym(VAR) %in% c(unlist(str_split(Config, " With "))) ~ "Group 1",
      TRUE ~ !!sym(VAR)
    ))

Is it a way to do it dynamically for each element of the vector and create the dedicated variable at each iteration ?

Thank you very much

1

There are 1 best solutions below

0
Umar On BEST ANSWER

is it ok

library(dplyr)
library(stringr)
library(purrr)
library(rlang)

# Sample data
Config <- c("4 With 1 + 6 With 2 + 8 With 3", "4 With 1 With 2 + 6 With 8")
mtcars <- data.frame(carb = c(4, 1, 2, 6, 8, 3)) # Sample data

# Function to create dynamic conditions
create_conditions <- function(config) {
  pairs <- str_split(config, "\\s*\\+\\s*")[[1]]
  
  conditions <- map2(pairs, 1:length(pairs), function(pair, group_num) {
    values <- as.numeric(str_extract_all(pair, "\\d+")[[1]])
    group_name <- paste("Group", group_num)
    condition <- expr(carb %in% c(!!!values) ~ !!group_name)
    return(condition)
  })
  
  return(conditions)
}

# Create a list of conditions for each element of Config
conditions_list <- map(Config, create_conditions)

# Create a dynamic column using case_when
mtcars <- mtcars %>%
  mutate(Merging = case_when(!!!set_names(unlist(conditions_list), NULL)))

# View the result
print(mtcars)