Time aggregate across years in R

Question

I'm trying to get the total rainfall that occurs between, for example, December and April of the following year. Another complexity is that the day can change, say Dec-15 to March-15.

I know that we can use aggregate or group_by if I don't have to go across the years. But I don't know how to solve this year's crossing problem.

Answer 1

Let say that 'x' is a Date class object or vector of Dates and 'ym' is the corresponding year/months, 'start_year' is the like years for the start of the season, and the 'n_season' is TRUE if the corresponding date is within Dec-Mar:

library(zoo)

x <- as.Date("2020-02-03")  # test data

ym <- as.yearmon(x)

start_year <- as.integer(ym - 3/12)

in_season <- cycle(ym) %in% c(12, 1:3)

You can subset to those dates in R programming, for which in_season is TRUE and then perform the aggregate() by the start_year.