In R, select values from another dataframe based on rows and columns, which are saved in dataframe

Question

I have two dataframes. The first, df1 contains countries and years. The second, df2, contains data that I want to include in df1 in a third column based on the match of the rows and columns of the respective values of df1.

df1
  country year
1       A 2008
2       B 2008
3       C 2009
4       F 2004
5       E 2006

df2
      country       2004       2005       2006       2007       2008       2009
    1       A 3,74972737 3,69814069  1,8119572  2,0058797  2,3728207 3,63424962
    2       B 3,62151043 1,54726382  -3,799075 1,92867306 2,92279764 0,68044437
    3       C 25,0489995 10,7724208 9,41065376 4,85433932 0,06592277 2,20000019
    4       F 4,78583195 5,04811878 3,46842543 3,78590254 4,19162568 4,01936553
    5       E 3,44897379 0,78317304 -2,2531746 2,74421327 1,79830266 0,23479692
    6       F 5,98651552 4,89339392 2,31922692 2,11685013 2,96275035 4,81028341
    7       G 5,65500512 7,29449815 2,96201437 5,37337313 6,62686519 6,45269876
    8       H 7,05863621 6,01378976 5,04512479 5,57180227 6,46438388 6,52143508
    9       I 7,67535068 3,63781612 -3,5861456 1,32402682 1,91501801 0,03094361

This is what I want to achieve:

  country year       gdp
1       A 2008  2.372821
2       B 2008  2.922798
3       C 2009  2.200000
4       F 2004  5.986516
5       E 2006 -2.253175

I am sure there is a very simple answer to this problem. How can I bring the data of df2 to df1?

I tried to use dplyr:mutate achieve it:

library(dplyr)

df1 <- mutate(df1, gdp = {
  df2[which(df2$country == country), 
      which(colnames(df2) == year)]})

However, the following Error message comes up

Error in which(colnames(df2) == year) : object 'year' not found

Answer 1

A solution using dplyr and tidyr . The key is to convert df2 to long format using gather . After that, we can conduct a merge operation with left_join . The last mutate call could be unnecessary if , in your data frame are all . . df3 is the final output.

library(dplyr)
library(tidyr)

df3 <- df1 %>%
  left_join(df2 %>% gather(year, gdp, -country, convert = TRUE), 
            by = c("country", "year")) %>%
  mutate(gdp = as.numeric(sub(",", "\\.", gdp)))
df3
#   country year       gdp
# 1       A 2008  2.372821
# 2       B 2008  2.922798
# 3       C 2009  2.200000
# 4       F 2004  4.785832
# 5       F 2004  5.986516
# 6       E 2006 -2.253175

DATA

df1 <- read.table(text = "country year
                  1       A 2008
                  2       B 2008
                  3       C 2009
                  4       F 2004
                  5       E 2006",
                  header = TRUE, stringsAsFactors = FALSE)

df2 <- read.table(text = "      country       2004       2005       2006       2007       2008       2009
    1       A 3,74972737 3,69814069  1,8119572  2,0058797  2,3728207 3,63424962
    2       B 3,62151043 1,54726382  -3,799075 1,92867306 2,92279764 0,68044437
    3       C 25,0489995 10,7724208 9,41065376 4,85433932 0,06592277 2,20000019
    4       F 4,78583195 5,04811878 3,46842543 3,78590254 4,19162568 4,01936553
    5       E 3,44897379 0,78317304 -2,2531746 2,74421327 1,79830266 0,23479692
    6       F 5,98651552 4,89339392 2,31922692 2,11685013 2,96275035 4,81028341
    7       G 5,65500512 7,29449815 2,96201437 5,37337313 6,62686519 6,45269876
    8       H 7,05863621 6,01378976 5,04512479 5,57180227 6,46438388 6,52143508
    9       I 7,67535068 3,63781612 -3,5861456 1,32402682 1,91501801 0,03094361",
                  header = TRUE, stringsAsFactors = FALSE)
names(df2) <- c("country", 2004:2009)