Conditional rolling mean (moving average) on irregular time series

Question

I have a group of data in the format:

ID    Minutes Value
xxxx  118     3 
xxxx  121     4 
xxxx  122     3 
yyyy  122     6 
xxxx  123     4 
yyyy  123              


        

Answer 1

An alternative approach that uses tidyr/dplyr instead of data.table and RcppRoll instead of zoo:

library(dplyr)
library(tidyr)
library(RcppRoll)

d %>% 
  group_by(ID) %>%
  # add rows for unosberved minutes
  complete(Minutes = full_seq(Minutes, 1)) %>%
  # RcppRoll::roll_mean() is written in C++ for speed 
  mutate(moving_mean = roll_mean(Value, 131, fill = NA, na.rm = TRUE)) %>%
  # keep only the rows that were originally observed
  filter(!is.na(Value))

data

d <- data_frame(
  ID = rep(1:3, each = 5),
  Minutes = rep(c(1, 30, 60, 120, 200), 3),
  Value = rpois(15, lambda = 10)
)