mutate multiple variables with different names in one function...?

Question

I am trying to replace update values for x, q, z using the recently available version of the variable indexed by date. In STATA, we can easily do it as a for loop (see sample code below). STATA

y refers to the date value of 20191125

local y 20191125

foreach v in attend child sibling{
replace `v'=`v'`y' if !missing(`v'`y')
}

Data

+----+----+---+-----------+-----------+-----------+
|attend  | child | sibling | attend20191125 | child20191125 | sibling20191125 |
+----+----+---+-----------+-----------+-----------+
| 1  | 2  | 3 |         6 |         8 |         0 |
| 1  | NA | 0 |         1 |         1 |         1 |
| NA | 0  | 1 |         5 |         4 |         2 |
+----+----+---+-----------+-----------+-----------+

Potential output:

+----+----+---+-----------+-----------+-----------+
|attend  | child | sibling | attend20191125 | child20191125 | sibling20191125 |
+----+----+---+-----------+-----------+-----------+
| 1  | 2  | 3 |         6 |         8 |         0 |
| 1  |1   | 0 |         1 |         1 |         1 |
| 5  | 0  | 1 |         5 |         4 |         2 |
+----+----+---+-----------+-----------+-----------+

I know how to replace NA values for one column from another column. How can I do the same using purrr dynamically? How can I tell R to replace values from a variable with the same name + date prefix? I have about 25 variables in my dataset that require this.

#Method 1: I can do it one variable at a time
df%<>%
mutate(attend=ifelse(is.na(attend)==T, attend20191125, attend),
       child=ifelse(is.na(child)==T, child20191125, child),
       sibling=ifelse(is.na(sibling)==T, sibling20191125, sibling))

#Method 2: using mutate_at (but not sure how I can dynamically refer to the date indexed variable?).

df%<>%
mutate_at(c("attend", "child", "sibling"), .=ifelse(is.na(.)==T, var20191125, .))

Yes, but I know how to replace missing values. My question is more focused on how I can ask R to replace the missing value from the new dated version? — brin, Nov 25 '19 at 17:30

score 1 · Accepted Answer · answered Nov 27 '19 at 01:46

This is one way:

get_replacement <- function(.) {
  name <- names(.)
  vals <- .[[1]]
  out <- ifelse(is.na(vals), df[[str_c(name, "20191125")]], vals)
  rlang::list2(!!name := out)  # dynamically set column name
}

df %>%
  lmap_at(vars(attend, child, sibling), get_replacement)

mutate multiple variables with different names in one function...?

1 Answers1