why write.csv is not updating the changes on test$number_s - r

Question

I am changing the values of a column of a data frame. Then, I am saving the file, supposedly with the changes, but not. What am I missing? Thanks,
test <- data.frame(name_s = c("x","y","z"), number_s = c(1,2,3))
lapply(1:length(test$number_s), function(x) {
test$number_s[x] <- test$number_s[[x]] + 1
})
write.csv(test,paste0("test ",format(Sys.time(),"%Y%m%d"),".csv"),
row.names = F)
that was oversimplified, the real deal is this one:
date_format_1 = "[0-9]-[:alpha:][:alpha:][:alpha:]"
date_format_2 = "[:alpha:][:alpha:][:alpha:]-[0-9][0-9]"
test <- data.frame(name_s = c("v","w","x","y","z"), event_text = c("Aug-89","7-May","9-Jun","4-Dec-2021","Feb-99"))
lapply(1:length(test$event_text), function(x) {
if (str_detect(test$event_text[[x]], paste0("\\b",date_format_1,"\\b")) == T){
test$event_text[x] <- paste0(str_sub(test$event_text[[x]],1,1), "/F",
which(month.abb %in% str_sub(
test$event_text[[x]], 3,5)))
} else if(str_detect(test$event_text[[x]], paste0("\\b", date_format_2,"\\b"))
== T) {
test$event_text[[x]] = paste0(which(month.abb %in% str_sub(
test$event_text[x],1,3)),"/F",str_sub(test$event_text[[x]],-2))
} else {
test$event_text[x] <- test$event_text[[x]]
}
})
write.csv(test,paste0("test ",format(Sys.time(),"%Y%m%d"),".csv"),
row.names = F)

Answer 1

Below I have written two calls to lapply that fix the issue you were having. The problem stems from the fact that R has scoped variables and so the value is changed within the function but the result is never returned or extracted from the function. As such I have demonstrated this by printing the dataframe after each of the lapply() calls below.
We can fix this in two ways. The first more correct version is to let lapply modify the exact vector directly by adding one to each value and returning x+1. (Note I have skipped curly braces and this will return the value from the next ppiece of code run, in this case x+1 alternatively you could write function(x) {return(x+1)} in that argument).
An alternate approach that will run slower but still use the indexing method is to use global assignment. <<- assigns the variable to the global scope/environment rather than the local scope of the function. (Note this code is run sequentially so the written call to this function is adding + 1 for the second time to the dataframe when shown below).
test <- data.frame(name_s = c("x","y","z"), number_s = c(1,2,3))
# Original Behaviour, doesn't work due to scoping issues
lapply(1:length(test$number_s), function(x) {
test$number_s[x] <- test$number_s[[x]] + 1
})
#> [[1]]
#> [1] 2
#>
#> [[2]]
#> [1] 3
#>
#> [[3]]
#> [1] 4
print(test)
#> name_s number_s
#> 1 x 1
#> 2 y 2
#> 3 z 3
# function that is syntactically and functionally correct
# instead of modifying the vector in the function scope the function returns the
# mutated vector which we then assign to the dataframe's vector
test$number_s <- lapply(test$number_s, function(x) x + 1)
print(test)
#> name_s number_s
#> 1 x 2
#> 2 y 3
#> 3 z 4
# function that is syntactically odd but functionally correct
# the function affects the values in the global scope, this works but is slower
# and is not best practice as it would be difficult to read
lapply(1:length(test$number_s), function(x) {
test$number_s[x] <<- test$number_s[[x]] + 1
})
#> [[1]]
#> [1] 3
#>
#> [[2]]
#> [1] 4
#>
#> [[3]]
#> [1] 5
print(test)
#> name_s number_s
#> 1 x 3
#> 2 y 4
#> 3 z 5
Created on 2021-07-23 by the reprex package (v2.0.0)