Calculated field based on Ranges in a second data frame in R - r

Question

I have found similar posts regarding this task, but all of which have a common ID joining the two tables.
I have one data frame which contains sale records (sales_df). For this example I have simplified the data table so that it contains only 5 records. I would like to create a new column in the sales_df that calculates what the fee would be given a sale price amount as defined in the fee table (pricing_fees). Please note that the number of actual pricing fee ranges that I have to account for are around 30, so writing this into a mutate statement is something that I would like to try and avoid.
The two data frames are coded as follows
sales_df <- data.frame(invoice_id = 1:5,
sale_price = c(100, 275, 350, 500, 675))
pricing_fees <- data.frame(min_range = c(0, 50, 100, 200, 300, 400, 500), # >=
max_range = c(50, 100, 200, 300, 400, 500, 1000), # <
buyer_fee = c(1, 1, 25, 50, 75, 110, 125))
In the end I would like the resulting sales_df to look something like this.
invoice_id sale_price buyer_fee
1 1 100 25
2 2 275 50
3 3 350 75
4 4 500 125
5 5 675 125
Thanks in advance

Answer 1

You can use findInterval function which is supposed to be efficient in splitting values over ranges (since it uses binary search) :
# build consecutive increasing ranges of fees
# (in order to use findInterval, since it works on ranges defined in a single vector)
pricing_fees <- pricing_fees[order(pricing_fees$min_range),]
consecFees <- data.frame(ranges=c(pricing_fees$min_range[1], pricing_fees$max_range),
fees=c(pricing_fees$buyer_fee,NA))
# consecFees now is :
#
# ranges fees
# 1 0 1 ---> it means for price in [0,50) -> 1
# 2 50 1 ---> it means for price in [50,100) -> 1
# 3 100 25 ---> it means for price in [100,200) -> 25
# 4 200 50 ... and so on
# 5 300 75
# 6 400 110
# 7 500 125
# 8 1000 NA ---> NA because for values >= 1000 we set NA
# add the column to sales_df using findInterval
sales_df$buyer_fee <- consecFees$fees[findInterval(sales_df$sale_price,consecFees$ranges)]
Result :
> sales_df
invoice_id sale_price buyer_fee
1 1 100 25
2 2 275 50
3 3 350 75
4 4 500 125
5 5 675 125

Answer 2

You can also use cut to "bin" sales_df$sale_price values and label bins with corresponding buyer_fee values.
# Make pricing_fee table with unique buyer_fee
brks <- do.call(rbind, by(pricing_fees, pricing_fees$buyer_fee, FUN = function(x)
data.frame(min_range = min(x$min_range), max_range = max(x$max_range), buyer_fee = unique(x$buyer_fee))))
sales_df$buyer_fee = as.numeric(as.character(cut(
sales_df$sale_price,
breaks = c(0, brks$max_range),
labels = brks$buyer_fee,
right = F)))
# invoice_id sale_price buyer_fee
#1 1 100 25
#2 2 275 50
#3 3 350 75
#4 4 500 125
#5 5 675 125