Nixtla's Open Source Time Series Ecosystem.

Nixtla Community

Currently trying to migrate from `fable`  to statsforecast in our production environment. However, I’m running into an issue with replication. Is it possible to `boxcox` transform prior to the forecasting step and back transform to produce the forecast mean rather than median? I’ve tried using the `scipy` python package, but it’s producing lower predictions than the R implementation. Any ideas / thoughts?

*Examples:*
*ETS* -&gt; `fable::ETS(fabletools::box_cox(qty + 1, lambda = 0.3), opt_crit = "mae")`
*ARIMA* -&gt; `fable::ARIMA((fabletools::box_cox(qty + 1, lambda = 0.4)) ~ PDQ(period = 13), stepwise = TRUE)`
*THETA* -&gt; `THETA(fabletools::box_cox(qty + 1, lambda = 0.1) ~ season(method = "additive"))`

Looks like it’s implemented in the `AutoARIMA` class, but still getting fairly low predictions: `AutoARIMA(stepwise=True,season_length=13,blambda=0.4,biasadj=0.4)`

Hi <@U03UTAG6GNP>, thanks for the question. We will have a look and come back to you with an answer.