How to combine two variables from different data frames in R? - r

Question

I'm new to R programming, so this question might be simple.
Anyway, I've tryed to find some answer to this specific thing I'm trying to do and didnt get it.
So, Im trying to import new data I've got to my old data.frame.
The problem is that this data has to substitute previous NA values in variables that already exist.
Also my data have different individuals (companys) in different periods (years), and my new data set only have the companys and years that was missing, plus some observation that I already had.
I tryied to simulate the problem with the data frames below:
Data frame with NAs:
df1 <- data.frame( company = c(rep("A",3), rep("B",3), rep("C",3)),
year = c(rep(2016:2018,each=1)),
income = c(95,87,93,NA,NA,58,102,80,NA),
debt = c(43,50,51,NA,37,37,53,NA,NA),
stringsAsFactors= F )
To search for new data, I created a data set with only the missing data, as my data had to many observations:
df_NA <- data.frame(df1[is.na(df1$income & df1$debt),])
So after searching, I was able to find the missing data, and now I have something like this:
df2 <- data.frame( company = c("A", "B" , "C" , "C"),
year = c(2018, 2016, 2017, 2018),
income = c(60,55, 80, 82),
debt = c(32,37, 53,48),
stringsAsFactors= F )
Now, I'm trying to get this data together, so I have the complete data.frame to work.
The problem is that I couldnt find a way to do it yet. I've tryed merge and join, indexing for company and year, but the variables that have the same name in both data.frame get duplicated and a suffix.
In my data I have much more observations and variables to fill, so I want to find a way I can do it with a command. Also this is going to happen again in the future, so it will be very helpfull.
I'm sorry if this was already answered. Thank you!

Answer 1

Here is an option using data.table:
library(data.table)
setDT(df1)
setDT(df2)
df1[df2, on=c("company", "year"), c('income', 'debt') := { list(i.income, i.debt)}]
# company year income debt
#1: A 2016 95 43
#2: A 2017 87 50
#3: A 2018 60 32
#4: B 2016 55 37
#5: B 2017 NA 37
#6: B 2018 58 37
#7: C 2016 102 53
#8: C 2017 80 53
#9: C 2018 82 48
Or another option using dplyr
library(dplyr)
full_join(df1, df2, by = c("year", "company")) %>%
mutate(
income = coalesce(income.x, income.y),
debt= coalesce(debt.x, debt.y),
) %>%
select(company, year, income, debt)