slight update to padded time series post http://weblog.bocoup.com/padding-time-series-with-r/

timefill.R

# if we have the data in a csv, we can make it more reproducible
# with dput(), specifically
# dput(read.csv("~/Desktop/ts.csv", as.is = TRUE), control = NULL)
# there are some peculiarities with list type objects and dput

unpadded.ts <- data.frame(list(time = c("2011/11/01", "2012/01/01", "2011/12/01", "2012/06/01"), observations = c(12, 320, 100, 7)))

# specifying the format will save you

unpadded.ts$time <- as.Date(unpadded.ts$time, format = "%Y/%m/%d")

# we can generate the padded ts pretty directly
# the dreaded recycling rule appears here. 

padded.ts <- data.frame(time = seq(min(unpadded.ts$time), max(unpadded.ts$time), by = "month"),
                        observations = 0)

# some subsetting and assignment and we're done!

padded.ts[padded.ts[, "time"] %in% unpadded.ts[, "time"], "observations"] <- unpadded.ts[, "observations"]