Question

I have an irregular time series that I’m working with that I’d like to convert to a regular one, but rather than the usual “data missing” behavior answered in other questions, I need to have the observation at each regularly spaced interval be the most recent observation, regardless of how long ago it was. I’ve written a function to do this, but with two loops it’s incredibly slow.

As an example, instead of having

> x <- zoo(c(1, 3, 6), c(1981, 1984, 1985))
> as.ts(x)
Time Series:
Start = 1981 
End = 1985 
Frequency = 1 
[1]  1 NA NA  3  6

I would like a result like this:

> as.ts(x)
Time Series:
Start = 1981 
End = 1985 
Frequency = 1 
[1]  1 1 1 3 6    

Answer 1

You can use na.locf from the zoo package.

y <- as.ts(x)
y <- na.locf(y)
y
# Time Series:
# Start = 1981 
# End = 1985 
# Frequency = 1 
# [1] 1 1 1 3 6