How to create unique vectors for large dataset - r

Question

I am trying to find the Atkinson Index measure for individual countries that spans over 11,000 observations. I have the decile measures for each specific observation which I can create an individual vector ex. c(d1, d2,...d10) for each single observation and compute the Atkinson Index but I am sure there is a quicker way to do this across 11,000 observations. Is there any possible way I can direct R to create a unique vectors across all 11,000 observations that use the deciles that are specific to each individual observation?
I am still rather new to coding in R, but I have tried to see if I can create some kind of loop that would return a vector pertaining to the deciles that corresponding with each individual observation.
id2 <- c(3.86, 5.29, 6.38, 7.32, 8.38, 9.35, 10.82, 12.47, 14.90, 21.22)
atkinson(id2, epsilon = 1)
[1] 0.1079504
Here is what I get when type:
dput(head(data))
structure(list(id = c(1, 2, 3, 4, 5, 6), country = c("Afghanistan",
"Albania", "Albania", "Albania", "Albania", "Albania"), c3 = c("AFG",
"ALB", "ALB", "ALB", "ALB", "ALB"), d1 = c(NA, 0, 3.49, 3.48,
3.73, 3.66), d2 = c(NA, 5.29, 4.86, 4.92, 5.14, 5.19), d3 = c(NA,
6.38, 5.84, 5.98, 6.09, 6.14), d4 = c(NA, 7.32, 6.74, 6.92, 6.98,
7.03), d5 = c(NA, 8.38, 7.65, 7.99, 7.91, 8.08), d6 = c(NA, 9.35,
8.84, 9.04, 8.92, 9.26), d7 = c(NA, 10.82, 10.23, 10.37, 10.3,
10.52), d8 = c(NA, 12.47, 11.98, 12.13, 11.93, 12.29), d9 = c(NA,
14.9, 14.93, 14.83, 14.54, 14.89), d10 = c(NA, 21.22, 25.44,
24.34, 24.46, 22.93)), row.names = c(NA, -6L), class = c("tbl_df",
"tbl", "data.frame"))
I can do this over 11,000 times but obviously that will take awhile, is there a way to construct R (a loop?) to do something along this lines for each individual observation?

Answer 1

Consider a row-wise calculation with apply to assign a new column to data frame. Underneath, as.vector() combines all decile points into a vector needed for atkinson().
data$atkinson_index <- apply(data[4:ncol(data)], MARGIN=1,
function(x) atkinson(as.vector(x), epsilon = 1)
)
data
Should NA pose a problem, wrap call in tryCatch
data$atkinson_index <- apply(data[4:ncol(data)], MARGIN=1,
function(x) tryCatch(atkinson(as.vector(x), epsilon = 1),
error = function(e) NA)
)
data