CODEWITHSUNDEEP
r
Mr. Buster
Mr. Buster

Reputation: 131

Create new variable based on the Look up table

I want to create a new variable on the data frame that uses a look up table. So I had df1 (dataframe) that has Amount and Term. And I need to create a new variable "Premium" that create its values using the look up table.

I tried the ifelse function but it's too tedious. Below is an illustration/example

df1 <- data.frame(Amount, Term)
df1
#   Amount Term
# 1   2500   23
# 2   3600   30
# 3   7000   45
# 4  12000   50
# 5  16000   38

And I need to create new variable the 'Premium' by using the Premium Lookup table below.

                  Term          
Amount           0-24 Mos  25-36 Mos 37-48 Mos 49-60 Mos
0 - 5,000         133      163       175       186
5,001 - 10,000    191      213       229       249
10,001 - 15,000   229      252       275       306
15,001 - 20,000   600      615       625       719
20,001 - 25,000   635      645       675       786

So the output for premium should be.

df1
#   Amount Term Premium
# 1   2500   23     133
# 2   3600   30     163
# 3   7000   45     229
# 4  12000   50     306
# 5  16000   38     625

Upvotes: 2

Views: 116

Answers (2)

thothal
thothal

Reputation: 20379

Data

df1 <- structure(list(Amount    = c(2500L, 3600L, 7000L, 12000L, 16000L), 
                      Term      = c(23L, 30L, 45L, 50L, 38L)), 
                 class     = "data.frame",
                 row.names = c(NA, -5L))

lkp  <- structure(c(133L, 191L, 229L, 600L, 635L, 
                    163L, 213L, 252L, 615L, 645L, 
                    175L, 229L, 275L, 625L, 675L, 
                    186L, 249L, 306L, 719L, 786L), 
                  .Dim      = 5:4, 
                  .Dimnames = list(Amount = c("0 - 5,000", "5,001 - 10,000",
                                              "10,001 - 15,000", "15,001 - 20,000", 
                                              "20,001 - 25,000"),
                                   Term   = c("0-24 Mos", "25-36 Mos", "37-48 Mos", 
                                              "49-60 Mos")))

Code

  1. Create first the upper limits for month and amount using regular expressions from the column and row names (you did not post your data in a reproducible way, so this regex may need adaptation based on your real lookup table structure):

    (month <- c(0, as.numeric(sub("\\d+-(\\d+) Mos$", 
                              "\\1", 
                              colnames(lkp)))))
# [1]  0 24 36 48 60

(amt   <- c(0, as.numeric(sub("^\\d+,*\\d* - (\\d+),(\\d+)$", 
                          "\\1\\2", 
                           rownames(lkp)))))
# [1]     0  5000 10000 15000 20000 25000

  2. Get the positions for each element of df1 using findInterval:

    (rows <- findInterval(df1$Amount, amt))
# [1] 1 1 2 3 4
(cols <- findInterval(df1$Term, month)) 
# [1] 1 2 3 4 3

  3. Use these indices to subset the lookup matrix:

    df1$Premium <- lkp[cbind(rows, cols)]
df1
#   Amount Term Premium
# 1   2500   23     133
# 2   3600   30     163
# 3   7000   45     229
# 4  12000   50     306
# 5  16000   38     625

Upvotes: 2

cephalopod
cephalopod

Reputation: 1906

To get to what you want you need to organise the table and categorise the data. I have provided a potential workflow to handle such situations. Hope this is helpful:

library(tidyverse)

df1 <- data.frame(
  Amount = c(2500L, 3600L, 7000L, 12000L, 16000L),
  Term = c(23L, 30L, 45L, 50L, 38L)
)

#~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~#
# functions for analysis ####
#~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~#

amount_tier_function <- function(x){

  case_when(x <= 5000   ~ "Tier_5000",
            x <= 10000  ~ "Tier_10000",
            x <= 15000  ~ "Tier_15000",
            x <= 20000  ~ "Tier_20000",
            TRUE        ~ "Tier_25000")
}


month_tier_function <- function(x){

  case_when(x <= 24   ~ "Tier_24",
            x <= 36   ~ "Tier_36",
            x <= 48   ~ "Tier_48",
            TRUE      ~ "Tier_60")
}

#~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~#
# Recut lookup table headings ####
#~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~#

lookup_df <- data.frame(stringsAsFactors=FALSE,
                amount_tier = c("Tier_5000", "Tier_10000", "Tier_15000", "Tier_20000",
                                "Tier_25000"),
                    Tier_24 = c(133L, 191L, 229L, 600L, 635L),
                    Tier_36 = c(163L, 213L, 252L, 615L, 645L),
                    Tier_48 = c(175L, 229L, 275L, 625L, 675L),
                    Tier_60 = c(186L, 249L, 306L, 719L, 786L)
             )

#~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~#
# Join everything together ####
#~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~#

lookup_df_tidy <- lookup_df %>% 
  gather(mth_tier, Premium, - amount_tier)


df1 %>%
  mutate(amount_tier = amount_tier_function(Amount),
         mth_tier    = month_tier_function(Term)) %>%
  left_join(., lookup_df_tidy) %>%
  select(-amount_tier, -mth_tier)

Upvotes: 0

Related Questions