[ANN] Durbyn.jl — Time Series Forecasting in Julia

Package Announcements
49

An initial benchmark for R forecast vs R smooth vs Julia Durbyn:

> packageVersion("forecast")
[1] ‘8.24.0’
> library(forecast)
> system.time({fit <- ets(AirPassengers, model = "ZZZ")})
   user  system elapsed 
   0.31    0.00    0.31 
> fc = forecast(fit, h = 12)
> autoplot(fc)

> packageVersion("smooth")
[1] ‘4.3.0’
> library(smooth)
> system.time({fit <- es(AirPassengers, model = "ZZZ")})
   user  system elapsed 
  0.198   0.000   0.198 
> fc = forecast(fit, h = 12)
> plot(fc)


using Durbyn
using Durbyn.ExponentialSmoothing

ap = air_passengers();
fit_ets = ets(ap, 12, "ZZZ")
@time fit_ets = ets(ap, 12, "ZZZ");
# 0.011112 seconds
fc_ets = forecast(fit_ets, h = 12)
plot(fc_ets)
2 Likes