I have a data.frame that doesn't account for leap year (ie all years are 365 days). I would like to repeat the last day value in February during the leap year. The DF in my code below has fake data set, I intentionally remove the leap day value in DF_NoLeapday. I would like to add a leap day value in DF_NoLeapday by repeating the value of the last day of February in a leap year (in our example it would Feb 28, 2004 value). I would rather like to have a general solution to apply this to many years data.
set.seed(55)
DF <- data.frame(date = seq(as.Date("2003-01-01"), to= as.Date("2005-12-31"), by="day"),
A = runif(1096, 0,10),
Z = runif(1096,5,15))
DF_NoLeapday <- DF[!(format(DF$date,"%m") == "02" & format(DF$date, "%d") == "29"), ,drop = FALSE]
We can use complete on the 'date' column which is already a Date class to expand the rows to fill in the missing dates
library(dplyr)
library(tidyr)
out <- DF_NoLeapday %>%
complete(date = seq(min(date), max(date), by = '1 day'))
dim(out)
#[1] 1096 3
out %>%
filter(date >= '2004-02-28', date <= '2004-03-01')
# A tibble: 3 x 3
# date A Z
# <date> <dbl> <dbl>
#1 2004-02-28 9.06 9.70
#2 2004-02-29 NA NA
#3 2004-03-01 5.30 7.35
By default, the other columns values are filled with NA, if we need to change it to a different value, it can be done within complete with fill
If we need the previous values, then use fill
out <- out %>%
fill(A, Z)
out %>%
filter(date >= '2004-02-28', date <= '2004-03-01')
# A tibble: 3 x 3
# date A Z
# <date> <dbl> <dbl>
#1 2004-02-28 9.06 9.70
#2 2004-02-29 9.06 9.70
#3 2004-03-01 5.30 7.35