Iterating through certain elements of nested list [R]-CodePudding

I have a large nested list. Its' structure is similar to the provided dummy data. I would like to loop through this list and apply a certain function only to the elements (vectors) named "seq" while ignoring the other categories. The elements of interest shall be subjected to the another function - a "function_of_interest" in the pseudocode below and the output shall be appended to the new list. Due to the size of my data, I would like to do it in parallel.

Here is the dummy input:

x <- list(one = list(one_1 = list(seq = 1:9, start = 1, end = 5), 
    one_2 = list(seq = 2:11, start = 2, end = 6), one_3 = list(
        seq = 3:12, start = 3, end = 7)), two = list(two_1 = list(
    seq = 1:13, start = 8, end = 222), two_2 = list(seq = 1:14, 
    start = 13, end = 54)))

And here is one of my attempts which failed:

#loop through the nested list

  for (gene in seq_along(genes_list)){
    for (segment in seq_along(genes_list[[gene]])){
      output_list <- c(output_list, foreach::foreach(segment) %dopar% function_of_interest(genes_list[[gene]][[segment]]))
    }
  }

Would be glad for help/guidance.

CodePudding user response：

How about this solution which calls lapply() on each element of x in parallel, and applies function_of_interest() to each element of the second-level nested list called seq. Note: this requires that each of those lists does in fact have an element named seq. You would need to add additional code to test whether this is the case for each list, if there is a possibility that some of them do not have an element called seq.

I defined a function_of_interest() to test your code.

function_of_interest <- function(vec) sum(vec)

output_list <- foreach(i = seq_along(x)) %dopar% {
  lapply(x[[i]], function(x_ij) function_of_interest(x_ij[['seq']]))
}

outputs:

[[1]]
[[1]]$one_1
[1] 45

[[1]]$one_2
[1] 65

[[1]]$one_3
[1] 75


[[2]]
[[2]]$two_1
[1] 91

[[2]]$two_2
[1] 105