Mapping a value based on two variables-CodePudding

I have a two data tables. One has company wide rating information, the other has price data at a debt-instrument level (company's usually have multiple debt-intsruments) for a historical date range.

The data table with the rating information looks as follows (simplified for illustration):

Company    Date         Rating
A          2016-02-01   AAA
A          2016-02-02   AA
B          2016-02-01   BBB
B          2016-02-02   A

The debt-instrument data frame looks as follows (simplified for illustration):

Company    Debt-Instrument     Date         Price
A          X1                  2016-02-01   100
A          X1                  2016-02-02   101
A          X2                  2016-02-01   98
A          X2                  2016-02-02   99
B          Y1                  2016-02-01   101
B          Y1                  2016-02-02   100
B          Y2                  2016-02-01   90
B          Y2                  2016-02-02   89

As you can see the rating information contains the company name and the date, which is also included in the debt-instrument table.

I would now like to add the rating information to the debt-instrument data table by matching the company name and the date.

The final data table should look as follows:

Company    Debt-Instrument     Date         Price  Rating
A          X1                  2016-02-01   100    AAA
A          X1                  2016-02-02   101    AA
A          X2                  2016-02-01   98     AAA
A          X2                  2016-02-02   99     AA
B          Y1                  2016-02-01   101    BBB
B          Y1                  2016-02-02   100    A
B          Y2                  2016-02-01   90     BBB
B          Y2                  2016-02-02   89     A

I know how to do a lookup on one single variable, however I do not know how to do it for two variables.

CodePudding user response：

You can use on = .() in data.table with multiple columns you would like the join like this:

df1 <- read.table(text = "Company    Date         Rating
A          2016-02-01   AAA
A          2016-02-02   AA
B          2016-02-01   BBB
B          2016-02-02   A", header = TRUE)

df2 <- read.table(text = "Company    Debt-Instrument     Date         Price
A          X1                  2016-02-01   100
A          X1                  2016-02-02   101
A          X2                  2016-02-01   98
A          X2                  2016-02-02   99
B          Y1                  2016-02-01   101
B          Y1                  2016-02-02   100
B          Y2                  2016-02-01   90
B          Y2                  2016-02-02   89", header = TRUE)

library(data.table)
setDT(df1)
setDT(df2)
df2[df1, on = .(Company, Date)]
#>    Company Debt.Instrument       Date Price Rating
#> 1:       A              X1 2016-02-01   100    AAA
#> 2:       A              X2 2016-02-01    98    AAA
#> 3:       A              X1 2016-02-02   101     AA
#> 4:       A              X2 2016-02-02    99     AA
#> 5:       B              Y1 2016-02-01   101    BBB
#> 6:       B              Y2 2016-02-01    90    BBB
#> 7:       B              Y1 2016-02-02   100      A
#> 8:       B              Y2 2016-02-02    89      A

^{Created on 2022-07-24 by the reprex package (v2.0.1)}

CodePudding user response：

With Base R merge function

merge(df2 , df1 , by = c("Company" , "Date"))

output

  Company       Date Debt.Instrument Price Rating
1       A 2016-02-01              X1   100    AAA
2       A 2016-02-01              X2    98    AAA
3       A 2016-02-02              X1   101     AA
4       A 2016-02-02              X2    99     AA
5       B 2016-02-01              Y1   101    BBB
6       B 2016-02-01              Y2    90    BBB
7       B 2016-02-02              Y1   100      A
8       B 2016-02-02              Y2    89      A