I'm not sure why but the following code some how creates a gap at year 0004 July 2nd:
library(tsibble)
library(gratis)
set.seed(2022)
sim1 <- arima_model(frequency = 7,
p = 1, # non-seasonal AR order
d = 0, # non-seasonal order of differencing
q = 0, # non-seasonal MA order
P = 1, # seasonal AR order
D = 0, # seasonal order of differencing
Q = 1, # seasonal MA order
constant = 0, # intercept
phi = c(0.8), # AR parameters
theta = c(), # MA parameters
Phi = c(-0.4), # seasonal AR parameters
Theta = c(0.8), # seasonal MA parameters
sigma = 0.5 # sd of noise
) %>%
generate(length = 2000, nseries = 1)
scan_gaps(sim1)
#> # A tsibble: 1 x 1 [1D]
#> index
#> <date>
#> 1 0004-07-02
This is actually a problem in tsibble::as_tsibble() due to the conversion of a numerical time index to an explicit date. Probably something for @earowang.
I'm not sure why but the following code some how creates a gap at year 0004 July 2nd:
Created on 2022-04-21 by the reprex package (v2.0.1)