madtoinou

Yes, your approach looks correct for multiple series training.

Hi @GuYith and @noahvand, Adding a section on the documentation about how to implement a custom model is on the roadmap but not a high priority at the moment. But...

Hi @dwolffram, Based on your example, the order is `[A1, B1, A2, B2]`. You can access each individual estimator using the helper method `model.get_multioutput_estimator(horizon, target_dim)`, to access the estimator for...

Well spotted, I noticed it as well and almost ready to submit a PR to fix this bug (and add docstring). I will also implement a method to get the...

Hi @ETTAN93, This is correct, even if the model does not rely on target lags, it'll just use its time index to generate the time index of the forecasts. So...

Hi @brianreinke95, In the documentation, we recommend using other libraries such as `Optuna` or `Ray-tune` to optimize hyper-parameters instead of `gridsearch()`. The tutorial you linked also applies to regression models...

Hi @dwolffram, `global_fit` refers to single versus multiple series, not the number of components in the series. Fitting a `Scaler` on all the components of a multivariate series is not...

Hi @markwkiehl, I just tried to reproduce the problem with the latest release (0.27.2) and everything went well. Could it be that you modified the cached file, hence the difference...

Hi @Allena101, This seems to be a purely modeling problem; the model are indeed seeing the entire dataset but optimizing their loss don't guarantee "coherence" of the forecast with the...

The `Diff()` transformer is reversible, you can indeed use inverse transform: ```python from darts.dataprocessing.transformers import Diff from darts.utils.timeseries_generation import linear_timeseries ts = linear_timeseries(length=100) tr_diff = Diff() ts_diff = tr_diff.fit_transform(ts) ts_inv_diff...