azul
azul
The SDK and the API handle the exogenous variables differently. The `SDK` differentiates between historical exogenous variables (used during finetuning, for example) and future ones. The historical variables are included...
Probably `tenacity` is not being invoked in distributed environments (namely `Dask`). See [this](https://github.com/Nixtla/nixtla/actions/runs/6794557008/job/18471163100#step:5:877).
Check that the target variable and exogenous variables are numerical before sending the request.
### Description it might be great to have a tutorial on using tbats and autotbats for multiple seasonalities based on the [mstl](https://nixtlaverse.nixtla.io/statsforecast/docs/tutorials/multipleseasonalities.html) one. ### Link _No response_
### Description Recently, we changed the statsforecast API to receive AnyDataframe instead of using a backend. The task consists of changing [this](https://nixtla.github.io/statsforecast/examples/prophet_spark_m5.html) tutorial and replacing the `backend` class with the...
### Description Transformations can often enhance the performance of certain types of models and data. We propose introducing a suite of composable transformations that can be applied to the time...
### Description To enhance the interpretability of our models, we could integrate a module or method to examine the importance of exogenous variables. The current challenge is that certain methods...
### Description Often, some pipelines require the ensembling of different statistical models. We can accomplish this by introducing a new module in the library named `statsforecast/ensembles.py`, which contains various methods...
### Description Only `AutoARIMA` and `MSTL-AutoARIMA` currently support exogenous regressors. A practical and effective approach is to first fit a model using the target variable and the exogenous regressors, and...