R：如何根据某些行的总和值删除行？

Question

Apologies if the question title is confusing - I wasn't sure how to frame it.

I have the following data frame:

df <- data.frame(
  comp_name = c("X", "A", "B", "C", "D", "Y", "E", "F", "G", "H", "Z", "J", "K", "L", "M"),
  parent_comp_name = c("X", "X", "X", "X", "X", "Y", "Y", "Y", "Y", "Y", "Z", "Z", "Z", "Z", "Z"),
  country = c("US", "US", "UK", "France", "Germany", "France", "US", "UK", "France", "Germany", "Germany", "US", "UK", "France", "Germany"),
  filing = c("Group", "Solo", "Solo", "Solo", "Solo", "Group", "Solo", "Solo", "Solo", "Solo", "Group", "Solo", "Solo", "Solo", "Solo"),
  profit = c(540, 100, 125, 150, 165, 495, 150, 110, 110, 125, 550, 130, 250, 95, 100)
)

data:

  comp_name parent_comp_name country filing profit
1          X                X      US  Group    540
2          A                X      US   Solo    100
3          B                X      UK   Solo    125
4          C                X  France   Solo    150
5          D                X Germany   Solo    165
6          Y                Y  France  Group    495
7          E                Y      US   Solo    150
8          F                Y      UK   Solo    110
9          G                Y  France   Solo    110
10         H                Y Germany   Solo    125
11         Z                Z Germany  Group    550
12         J                Z      US   Solo    130
13         K                Z      UK   Solo    250
14         L                Z  France   Solo     95
15         M                Z Germany   Solo    100

This data frame is a simplified version of the actual data I am working with.

I want to write a script which checks: For a given parent company (say X), if the sum of profits for all the solo parent_company X filings is equal to the group filing profit, delete the solo rows.

I want the output table to look like this:

  comp_name parent_comp_name country filing profit
1         X                X      US  Group    540
2         Y                Y  France  Group    495
3         Z                Z Germany  Group    550
4         J                Z      US   Solo    130
5         K                Z      UK   Solo    250
6         L                Z  France   Solo     95
7         M                Z Germany   Solo    100

Here you can see that the solo filings for parent_comp_name X and Y have been removed, as their profits summed up to the respective group total profit. However, rows for company Z were not removed as the sum of solo profits did not add up to the group.

I am relatively new to R and do not know how to go about getting started with this. Any help would be greatly appreciated. Thanks!

Answer 1

Basic idea: keep all rows with filing == "Group" and find which rows with filing == "Solo" to keep.

library(tidyverse)

keep <- df %>%
  group_by(parent_comp_name, filing) %>%
  summarise(s = sum(profit)) %>%
  ungroup() %>%
  pivot_wider(names_from = filing, values_from = s) %>%
  filter(Group != Solo)  %>%
  pluck("parent_comp_name") %>%
  as.character()

df %>%
  filter(filing == "Group" | parent_comp_name %in% keep) 

  comp_name parent_comp_name country filing profit
1         X                X      US  Group    540
2         Y                Y  France  Group    495
3         Z                Z Germany  Group    550
4         J                Z      US   Solo    130
5         K                Z      UK   Solo    250
6         L                Z  France   Solo     95
7         M                Z Germany   Solo    100

Answer 2

Here's a way with dplyr -

library(dplyr)

df %>%
  group_by(parent_comp_name) %>%
  filter(if(sum(profit[filing == 'Solo']) == sum(profit[filing != 'Solo'])) 
         filing != 'Solo' else TRUE) %>% 
  ungroup

# comp_name parent_comp_name country filing profit
#  <chr>     <chr>            <chr>   <chr>   <dbl>
#1 X         X                US      Group     540
#2 Y         Y                France  Group     495
#3 Z         Z                Germany Group     550
#4 J         Z                US      Solo      130
#5 K         Z                UK      Solo      250
#6 L         Z                France  Solo       95
#7 M         Z                Germany Solo      100

For a parent_comp_name , if sum of profit for filling = 'Solo' is equal to sum of profit for non-solo drop rows where filing = 'Solo' .

Answer 3

Slightly longer version of Ronak's code but another approach:

library(dplyr)
df %>% group_by(parent_comp_name) %>% 
   mutate(grp_profit = sum(profit[filing == 'Group']), solo_profit = sum(profit[filing == 'Solo'])) %>% 
     filter(if(grp_profit == solo_profit) filing == 'Group' else TRUE) %>% select(-c(grp_profit,solo_profit))
# A tibble: 7 x 5
# Groups:   parent_comp_name [3]
  comp_name parent_comp_name country filing profit
  <chr>     <chr>            <chr>   <chr>   <dbl>
1 X         X                US      Group     540
2 Y         Y                France  Group     495
3 Z         Z                Germany Group     550
4 J         Z                US      Solo      130
5 K         Z                UK      Solo      250
6 L         Z                France  Solo       95
7 M         Z                Germany Solo      100