R: reconstruct manager-employee data - r

Question

I have a list of employee information with employee id and direct line manager id. I want to rearrange the data so it will list all level of managers for each employee.
I want to create a loop to find line managers repeatedly.
Here is the code to create a sample dataset.
employee_id = seq(1:10)
manager_id =c(1,1,2,3,4,2,3,1,4,5)
hr=data.frame(employee_id,manager_id)
Here is what I expect:
Using employee_id 4 as an example
employee_id managerL1 managerL2 managerL3
4 3 2 1
I should also mention that this is a simplified example. In the real data that I'm working with, manager and employee ids are not sequential. They are some random numbers with prefixs. The id itself don't have any information on managerial levels. The level is purely driven by data.

Answer 1

Seems like this requires an iterative solution.
Start with the level 1 managers of our employees. The row index of the employee who is the manager of each employee is
i <- 1
idx = match(hr$manager_id, hr$employee_id)
The manager's manager is hr$manager_id[idx], and we can use the same match() approach iteratively. Record and repeat until there is just a single employee as manager
repeat {
idx = match(hr$manager_id[idx], hr$employee_id)
hr[[paste0("manager_", i)]] = hr$employee_id[idx]
if (length(unique(idx)) == 1)
break
i <- i + 1
}
A variant might allow for one or more top-level managers by using NA as their manager, and stopping appropriately
hr$employee_id[1] = NA # the boss; there could be several top-level managers...
i <- 1
idx = match(hr$manager_id, hr$employee_id)
repeat {
idx = match(hr$manager_id[idx], hr$employee_id)
hr[[paste0("manager_", i)]] = hr$employee_id[idx]
if (all(is.na(idx)))
break
i <- i + 1
}

Answer 2

Here is an option with tidyverse
library(tidyverse)
hr %>%
uncount(manager_id, .remove = FALSE) %>%
group_by(employee_id) %>%
mutate(new_id = row_number(), nm1 = str_c('manager_', new_id)) %>%
spread(nm1,new_id)
# A tibble: 10 x 7
# Groups: employee_id [10]
# employee_id manager_id manager_1 manager_2 manager_3 manager_4 manager_5
# <int> <dbl> <int> <int> <int> <int> <int>
# 1 1 1 1 NA NA NA NA
# 2 2 1 1 NA NA NA NA
# 3 3 2 1 2 NA NA NA
# 4 4 3 1 2 3 NA NA
# 5 5 4 1 2 3 4 NA
# 6 6 2 1 2 NA NA NA
# 7 7 3 1 2 3 NA NA
# 8 8 1 1 NA NA NA NA
# 9 9 4 1 2 3 4 NA
#10 10 5 1 2 3 4 5
Or with map and spread
hr %>%
mutate(new_id = map(manager_id, seq)) %>%
unnest %>%
mutate(nm1 = str_c('manager_', new_id)) %>%
spread(nm1, new_id)