在 R 中的嵌套列表中计算

Question

How do I calculate the sum of the count per city in nested lists and store it in a new tibble using map?

The output of the nested lists looks like this:

City:'s-Hertogenbosch`

$`'s-Hertogenbosch`$`5392`

$`'s-Hertogenbosch`$`5392`$men

 A tibble: 20 x 2

   age              count
   <chr>            <dbl>
 1 0 tot 5 jaar        20
 2 5 tot 10 jaar       40
 3 10 tot 15 jaar      45
 4 15 tot 20 jaar      20
 5 20 tot 25 jaar      25


$`'s-Hertogenbosch`$`5392`$women
 A tibble: 20 x 2
   age              count
   <chr>            <dbl>
 1 0 tot 5 jaar        15
 2 5 tot 10 jaar       30
 3 10 tot 15 jaar      35
 4 15 tot 20 jaar      30
 5 20 tot 25 jaar      15



City:'Aa en Hunze`
$`Aa en Hunze`
$`Aa en Hunze`$`9443`
$`Aa en Hunze`$`9443`$men
 A tibble: 20 x 2
   age              count
   <chr>            <dbl>
 1 0 tot 5 jaar         0
 2 5 tot 10 jaar       10
 3 10 tot 15 jaar       5
 4 15 tot 20 jaar       5
 5 20 tot 25 jaar       5


$`Aa en Hunze`$`9443`$women
 A tibble: 20 x 2
   age              count
   <chr>            <dbl>
 1 0 tot 5 jaar         5
 2 5 tot 10 jaar        5
 3 10 tot 15 jaar       5
 4 15 tot 20 jaar      10
 5 20 tot 25 jaar       5

I know how to store 1 specific postal code in a tibble:

sum of count <- tibble(sum(data[[1]][[1]][[1]]['count']))

But i can't figure out how to apply this for all lists in R using map:

Get_count_per_postal <- map_int(data, function(x){
    tibble(municipality = names(data), count_inhibitants = sum(data[[?]][[?]][[?]]['count']))
  })

I'm having a hard time accessing the nested lists and apply a calculation, it would be great if someone set me on right path.

Kr, Quido

Answer 1

It is not clear clear about the structure correctly. Perhaps, a recursive option would work - use rrapply to change the structure by melt ing the nested list , then filter the column 'L4' having only 'count', unnest the list column 'value', do a group_by sum (if we don't want to include 'L3' ie sex as grouping, remove it to get the overall count per 'municipality'

library(rrapply)
library(dplyr)
library(tidyr)
rrapply(data, how = 'melt') %>%
    filter(L4 == 'count') %>% 
    unnest(value) %>%
    group_by(municipality = L1, sex = L3) %>%
    summarise(count_inhabitants = sum(value, na.rm = TRUE), .groups = 'drop')
# A tibble: 4 × 3
  municipality          sex   count_inhabitants
  <chr>                 <chr>             <dbl>
1 City:'Aa en Hunze     men                  25
2 City:'Aa en Hunze     women                30
3 City:'s-Hertogenbosch men                 150
4 City:'s-Hertogenbosch women               125

data

data <- list(`City:'s-Hertogenbosch` = list(`'s-Hertogenbosch$5392` = list(
    men = structure(list(age = c("0 tot 5 jaar", "5 tot 10 jaar", 
    "10 tot 15 jaar", "15 tot 20 jaar", "20 tot 25 jaar"), count = c(20, 
    40, 45, 20, 25)), class = c("tbl_df", "tbl", "data.frame"
    ), row.names = c(NA, -5L)), women = structure(list(age = c("0 tot 5 jaar", 
    "5 tot 10 jaar", "10 tot 15 jaar", "15 tot 20 jaar", "20 tot 25 jaar"
    ), count = c(15, 30, 35, 30, 15)), class = c("tbl_df", "tbl", 
    "data.frame"), row.names = c(NA, -5L)))), `City:'Aa en Hunze` = list(
    `'Aa en Hunze$9443` = list(men = structure(list(age = c("0 tot 5 jaar", 
    "5 tot 10 jaar", "10 tot 15 jaar", "15 tot 20 jaar", "20 tot 25 jaar"
    ), count = c(0, 10, 5, 5, 5)), class = c("tbl_df", "tbl", 
    "data.frame"), row.names = c(NA, -5L)), women = structure(list(
        age = c("0 tot 5 jaar", "5 tot 10 jaar", "10 tot 15 jaar", 
        "15 tot 20 jaar", "20 tot 25 jaar"), count = c(5, 5, 
        5, 10, 5)), class = c("tbl_df", "tbl", "data.frame"), row.names = c(NA, 
    -5L)))))