Nested loops in R that create new variable names and lags - r

Question

I am new to R, but experienced in Stata. To learn R, I am tracking Covid-19 infections. That requires creating seven-day trailing averages, and I do so with the following loop.
for (mylag in c(1:7)) {
data <- data %>% group_by(state) %>% mutate(!!paste0("deathIncrease", "_", mylag) := lag(deathIncrease, mylag)) %>% ungroup()
}
This works, but then I want to run the same code, not just for deaths, but also for cases. So I tried the following.
var_list <- c("deathIncrease", "positiveIncrease")
for (var in var_list) {
for (mylag in c(1:7)) {
var <- enquo(var)
varname <- enquo( paste0(quo_name(var), "_", mylag) )
data <- data %>% group_by(state) %>% mutate(!!varname := lag(!!var, mylag)) %>% ungroup()
}
}
But that leads to the error arg must be a symbol. Any help would be much appreciated. In Stata, loops are simpler. Is there no package that gets R to automatically fill in the looping variables everywhere, like so: {{ var }}?
Edit: here is a minimal working example. The first way to create lags works, but only for var1. The second nested loop does not.
df <- tribble(
~group_var, ~var1, ~var2,
"A", 1, 10,
"A", 2, 11,
"A", 3, 12,
"B", 1, 10,
"B", 2, 11,
"B", 3, 12)
for (mylag in c(1:2)) {
df <- df %>% group_by(group_var) %>% mutate(!!paste0("var1", "_lag", mylag) := lag(var1, mylag)) %>% ungroup()
}
## Another loop
var_list <- c("var1", "var2")
for (myvar in var_list) {
for (mylag in c(1:2)) {
myvar <- enquo(myvar)
varname <- enquo( paste0(quo_name(myvar), "_", mylag) )
data <- data %>% group_by(state) %>% mutate(!!varname := lag(!!myvar, mylag)) %>% ungroup()
}
}

Answer 1

You can use the function get(), like lag(get(myvar), mylag), to point the specific column the string myvar is referred to:
for(mylag in 1:7){
for(myvar in c('deathIncrease', 'positiveIncrease')){
data <- data %>%
group_by(state) %>%
mutate(
!!paste0(myvar, '_', mylag) := lag(get(myvar), mylag)
) %>%
ungroup()
}
}

Answer 2

My first solution contained a function, that did not respect grouped data. I wanted to look at that anyways, so i spend a bit of time to respect grouped data as well.
This is my solution now, it works as expected on grouped data, but it feels a bit hacky tbh.
add_lag <- function(.data, column, days) {
group <- unlist(groups(.data))
if(is.null(group)){
new <- mapply(function(x, y) {
lag(x, y)
}, x = .data[column], y = sort(rep(days, length(column))))
if(is.null(dim(new))){
new <- t(new)
}
new <- as.data.frame(new, stringsAsFactors = F)
names(new) <- paste0(column, "_", sort(rep(days, length(column))))
new <- as_tibble(new) %>%
select(sort(names(new)))
mutate(.data, !!!new)
} else {
tmp <- .data %>%
nest()
tmp$data <- lapply(tmp$data, function(x,y){
x %>%
add_lag(column,y)
}, y = days)
tmp %>% unnest(c(data))
}
}
df <- tribble(
~group_var, ~var1, ~var2,
"A", 1, 10,
"A", 2, 11,
"A", 3, 12,
"B", 1, 10,
"B", 2, 11,
"B", 3, 12)
df %>%
group_by(group_var) %>%
add_lag("var1", 1:2)
# A tibble: 6 x 5
# Groups: group_var [2]
group_var var1 var2 var1_1 var1_2
<chr> <dbl> <dbl> <dbl> <dbl>
1 A 1 10 NA NA
2 A 2 11 1 NA
3 A 3 12 2 1
4 B 1 10 NA NA
5 B 2 11 1 NA
6 B 3 12 2 1