R：根据子集数据帧的总和来子集数据帧条件

Question

I have a data frame manipulation question.

I would like to find the subset of data frame "data1" which sum of each col equal to another data frame "data2".

Here is my code:

AA<-c(2,3,1,4,9)
BB<-c(5,13,9,1,2)

A1<-c(5)
B1<-c(18)

data1<-data.frame(AA,BB)
data2<-data.frame(A1,B1)

library(dplyr)
subset(data1, ((sum(AA) ==data2$A1 )  &&  (sum(BB) ==data2$B1 ) ) )

I am wondering if any other algorithm would help?

Thanks!

Answer 1

This solution only considers the scenario that you want to calculate the sum from any two rows. If you want to test other row numbers, you will need to create those combinations by changing the numbers in the combn function. final_data is the final output. If there are multiple matches, you may want to keep the final_data as a list.

# Prepare example datasets
AA<-c(2,3,1,4,9)
BB<-c(5,13,9,1,2)

A1<-c(5)
B1<-c(18)

data1<-data.frame(AA,BB)
data2<-data.frame(A1,B1)

# Load packages
library(tidyverse)

# Use combn to find out all the combination of row number
row_indices <- as.data.frame(t(combn(1:nrow(data1), 2)))

# Prepare a list of data frame. Each data frame is one row from row_indices
row_list <- row_indices %>%
  rowid_to_column() %>%
  split(f = .$rowid)

# Based on row_list to subset data1
sub_list <- map(row_list, function(dt){
  temp_data <- data1 %>% filter(row_number() %in% c(dt$V1, dt$V2))
  return(temp_data)
})

# Calcualte the sum of each data frame in sub_list
sub_list2 <- map(sub_list, function(dt){
  dt2 <- dt %>% 
    summarise_all(funs(sum(.))) %>%
    setNames(c("A1", "B1"))
  return(dt2)
})

# Compare each data frame in sub_list2 with data2
# Find the one that is the same and store the logical results in result_indices
result_indices <- map_lgl(sub_list2, function(dt) setequal(dt, data2))

# Get the final output
final_data <- sub_list[result_indices][[1]]

final_data
  AA BB
1  2  5
2  3 13