Apply function with different sources in R-CodePudding

I'm having trouble using the apply function and repeatedly get an error about different sources.

I believe that both sources are data.frames, so I can't work out why it doesn't like the apply inputs.

I would like to get to the output_desired without the need to run them individually. I appreciate I could likely do this with a loop, but am trying to learn some need tricks.

TLDR; I am trying to group the input_s into an object, and then run that dataframe through the function(result) and rbind the outputs.

library(dplyr)
library(tidyr)

## Inputs ##

input_1 = structure(list(V1 = c("Team_2022", "Team_2022", "Team_2022"), V2 = c("Frank", "Mary", "John"), V3 = c("Sydney", "Sydney", "Sydney"), V4 = c(55, 76, 14)), row.names = c(NA, -3L), class = c("data.table", "data.frame"))
input_2 = structure(list(V1 = c("Team_2023", "Team_2023", "Team_2023"), V2 = c("Bill", "Mary", "John"), V3 = c("Sydney", "Sydney", "Sydney"), V4 = c(113, 23, 10)), row.names = c(NA, -3L), class = c("data.table", "data.frame"))
input_3 = structure(list(V1 = c("Team_2024", "Team_2024", "Team_2024"), V2 = c("Frank", "Mary", "Bill"), V3 = c("Sydney", "Sydney", "Sydney"), V4 = c(7, 19, 52)), row.names = c(NA, -3L), class = c("data.table", "data.frame"))
input_4 = structure(list(V1 = c("Team_2025", "Team_2025", "Team_2025"), V2 = c("Frank", "Mary", "John"), V3 = c("Sydney", "Sydney", "Sydney"), V4 = c(46, 44, 88)), row.names = c(NA, -3L), class = c("data.table", "data.frame"))

## Teams ##

teams = structure(list(V1 = c("team1", "team2", "team3"), V2 = c("Mary   Frank","Mary   John", "Mary   Bill")), class = "data.frame", row.names = c(NA, -3L))

## Group the inputs into one ##

all_objects = ls()
input_objects = grep("^input", all_objects, value = T)
input_test = as.data.frame(input_obj)

## Function ##

result = function(input, teams) {
  data = teams %>%
  separate_rows(V2) %>%
  left_join(input, by = c("V2" = "V2")) %>%
  replace_na(list(V4 = 0)) %>%
  group_by(V1.x) %>% fill(V1.y, V3) %>%
  summarize(V1.y = first(V1.y),
            V2 = paste(V2, collapse = "   "),
            V3 = first(V3),
            V4 = sum(V4))
  return(data)
}

## Outputs individually ##

output_1 = result(input_1, teams)
output_2 = result(input_2, teams)
output_3 = result(input_3, teams)
output_4 = result(input_4, teams)

## Join outputs ##

output_desired = rbind(output_1, output_2, output_3, output_4)

## Failed apply ##

output_apply = apply(input_test, 1, function(x) {
      result(x, teams)
}) %>% do.call("rbind", .)

CodePudding user response：

Using lapply and dplyr::bind_rows you could do:

Note: I also fixed the creation of your input_test list for which I use lapply and get.

library(dplyr)

all_objects <- ls()
input_objects <- grep("^input_\\d", all_objects, value = T)
input_test <- lapply(input_objects, get)

input_test %>%
  lapply(result, teams)  %>%
  bind_rows()
#> # A tibble: 12 × 5
#>    V1.x  V1.y      V2           V3        V4
#>    <chr> <chr>     <chr>        <chr>  <dbl>
#>  1 team1 Team_2022 Mary   Frank Sydney   131
#>  2 team2 Team_2022 Mary   John  Sydney    90
#>  3 team3 Team_2022 Mary   Bill  Sydney    76
#>  4 team1 Team_2023 Mary   Frank Sydney    23
#>  5 team2 Team_2023 Mary   John  Sydney    33
#>  6 team3 Team_2023 Mary   Bill  Sydney   136
#>  7 team1 Team_2024 Mary   Frank Sydney    26
#>  8 team2 Team_2024 Mary   John  Sydney    19
#>  9 team3 Team_2024 Mary   Bill  Sydney    71
#> 10 team1 Team_2025 Mary   Frank Sydney    90
#> 11 team2 Team_2025 Mary   John  Sydney   132
#> 12 team3 Team_2025 Mary   Bill  Sydney    44

Or using purrr::map_df:

purrr::map_df(input_test, result, teams)