CODEWITHSUNDEEP
rdata.tableinterpolation
Amitai
Amitai

Reputation: 891

R data.table - how to interpolate missing values over FIXED rows in multiple columns

a = data.table(c(2,NA,3), c(5,NA,1))

When I try to interpolate over the missing lines

a[, approx(x = 1:.N, y = .SD, xout = which(is.na(.SD))), .SDcols = 1:2]

gives the following error:

Error in xy.coords(x, y) : 'x' and 'y' lengths differ

I wish to get the following:

> a
    V1 V2
1: 2.0  5
2: 2.5  3
3: 3.0  1

Upvotes: 3

Views: 761

Answers (1)

Arun
Arun

Reputation: 118799

It seems like x and y (first two arguments) should be numeric vectors. You'll need to loop through each column.. Here I use set() along with a for-loop to update the original data.table by reference.

len = 1:nrow(a)
for (col in names(a)) {
    nas = which(is.na(a[[col]]))
    set(a, i=nas, j=col, value=approx(len, a[[col]], xout=nas)$y)
}
#     V1 V2
# 1: 2.0  5
# 2: 2.5  3
# 3: 3.0  1

Upvotes: 4

Related Questions