Dynamically adding up column variables-CodePudding

I need help in dynamically adding up variables and evaluating them inside a dplyr pipe.

For example I have a dataset and I have to go:

test_variable_construction %>%
  mutate(Total_Utilisation_Sum_3m = Total_Utilisation Lag1_Total_Utilisation Lag2_Total_Utilisation)

and

test_variable_construction %>%
  mutate(Total_Utilisation_Sum_6m = Total_Utilisation   Lag1_Total_Utilisation   Lag2_Total_Utilisation   Lag3_Total_Utilisation   Lag4_Total_Utilisation   Lag5_Total_Utilisation)

so I want to create a string as such:

paste(iterates,collapse=" ")

"Total_Utilisation Lag1_Total_Utilisation Lag2_Total_Utilisation"

and pass this into the dplyr pipe:

test_variable_construction %>%
mutate(Total_Utilisation_Sum_3m := eval(!!paste(iterates,collapse=" "))

Unfortunately, that doesn't work -- does anyone know the right syntax to for this?

CodePudding user response：

One option to achieve your desired result would be to use rowwise like so:

library(dplyr)

test_df <- data.frame(Total = 1:5)
test_df[paste0("Lag", 1:2, "_Total")] <- lapply(1:2, function(x) 1:5)

iterates <- c("Total", paste0("Lag", 1:2, "_Total"))

test_df %>% 
  rowwise() %>% 
  mutate(Total_Sum_3m = sum(c_across(all_of(iterates)))) %>% 
  ungroup()
#> # A tibble: 5 × 4
#>   Total Lag1_Total Lag2_Total Total_Sum_3m
#>   <int>      <int>      <int>        <int>
#> 1     1          1          1            3
#> 2     2          2          2            6
#> 3     3          3          3            9
#> 4     4          4          4           12
#> 5     5          5          5           15

Or using rowSums you could do:

test_df %>% 
  mutate(Total_Sum_3m = rowSums(.[iterates]))
#>   Total Lag1_Total Lag2_Total Total_Sum_3m
#> 1     1          1          1            3
#> 2     2          2          2            6
#> 3     3          3          3            9
#> 4     4          4          4           12
#> 5     5          5          5           15

Note: Using eval you could achieve your result via

test_df %>%
  mutate(Total_Sum_3m = eval(parse(text = paste(iterates, collapse = " "))))

However, in general using eval(parse = ... is not recommended. See e.g. this post for a collection of reasons to avoid eval(parse = ....

CodePudding user response：

So the parse_expr from rlang worked for me.

For example, I had to go:

test_variable_construction %>%
mutate(Total_Utilisation_Sum_3m = !!parse_expr(paste(iterates), collapse = "   "))