Sum values in a comma-separated string - r

Question

I have a dataset that has a column called QTY in which most of the values are already summed, but a few are several integers separated by commas. How can I replace those rows with the sums of the values?
I have:
ID Name QTY
1 Abc 2
2 Bac 3
3 Cba 2, 4, 5, 8
4 Bcb 4, 1
Desired result:
ID Name QTY
1 Abc 2
2 Bac 3
3 Cba 19
4 Bcb 5
I've tried messing around with for loops a bit and using ifelse(), but I can't quite figure it out.

Answer 1

This looks a bit ugly but should work. Assuming column QTY is a character -
your_df$QTY_new <- sapply(strsplit(your_df$QTY, ", "), function(x) sum(as.numeric(x)))

Answer 2

Using for loops should be this way:
data <- data.table(ID = 1:4,
Name = c("Abc", "Bac", "Cba", "Bcb"),
QTY = c("2", "3", "2, 4, 5, 8", "4, 1"),
QTY2 = numeric(4))
for(i in 1:nrow(data)){
data$QTY2[i] <- sum(as.numeric(unlist(strsplit(as.character(data$QTY[i]), ', '))))
}
and the resulting DF is:
ID Name QTY QTY2
1: 1 Abc 2 2
2: 2 Bac 3 3
3: 3 Cba 2, 4, 5, 8 19
4: 4 Bcb 4, 1 5

Answer 3

I made a function for solving your question. But let me explain how it works:
sumInRow = function(row_value, split = ",") {
# 1. split the values
row_value = strsplit(row_value, split = split)
# 2. Convert them to numeric and sum
row_sum = sapply(row_value, function(single_row) {
single_row = as.numeric(single_row)
return(sum(single_row))
})
return(row_sum)
}
The row_value, by default, will be a character because of the comma.
Then for each value we need to split them:
row_value = strsplit(row_value, split = split)
But it will return a list contain the split for all element in row_value, don't worry we'll use it later.
row_sum = sapply(row_value, function(single_row) {
single_row = as.numeric(single_row)
return(sum(single_row))
})
Sapply function works as an interator, for each element of the list we'll use the following function: convert to numeric and return the sum of them.
[EDIT_1]
To use if you have to call:
sumInRow(<your data frame>$QYT)
I hope it helps you.

Answer 4

Here is one option with tidyverse, We split the 'QTY' column by the delimiter , to expand the rows (separate_rows), grouped by 'ID', 'Name', get the sum of the 'QTY'
library(tidyverse)
df1 %>%
separate_rows(QTY, convert = TRUE) %>%
group_by(ID, Name) %>%
summarise(QTY = sum(QTY))
# A tibble: 4 x 3
# Groups: ID [4]
# ID Name QTY
# <int> <chr> <int>
#1 1 Abc 2
#2 2 Bac 3
#3 3 Cba 19
#4 4 Bcb 5
data
df1 <- structure(list(ID = 1:4, Name = c("Abc", "Bac", "Cba", "Bcb"),
QTY = c("2", "3", "2, 4, 5, 8", "4, 1")), class = "data.frame", row.names = c(NA,
-4L))