-
If there is a sharp increase in the number of events (e.g., hospitalizations), the TSENS model will fail. A few recent examples showing locations for influenza hospitalizations where TSENS failed (the…
-
I feel very limited that SSA is the only method available for time series modeling. Of course, other parametric and non-parametric models are capable of forecasting, but a parametric approach like ARI…
-
In #22 I'm working on using auto-ARIMA to select better models for each variable, but it is time consuming. Once the model selection process has happened though, you can use `refit()` or `stream()` f…
-
#### Describe the bug
SARMIXResults `remove_data` does not allow for persisted model to still execute forecast method. As an example, my current model joblib is 1.9Gb and reduces to 200Mb with `rem…
-
**Describe the bug**
Calling add_holidays on timeseries makes multivariate series.
**Expected behavior**
Calling models that can handle holidays (eg. Prophet, ARIMA with exogenous series) would a…
-
Hi. Your code for MA linear regression uses residual for y_train
But the following …
-
**Problem:** When applying the Ljung-Box test to residuals from ARMA(p, q) models with k lags, Ljung & Box (1978) and also Box & Pierce (1970) say that the degrees of freedom need to be adjusted by p+…
-
While `AutoARIMA` from `pmdarima` is slow, it shows the model hyperparameters in a convenient manner. How can I get the same for `statsforecast` `AutoARIMA`?
A consistent interface with `pmdarima` …
-
I want to forward validate several auto AutoARIMA models wrapped in a StatsForecast object (multiple time series).
Is there a way to freeze once found optimal auto arima parameters to re use them (…
-
**Describe the bug**
`ForecastX` returns weird predictions, when I recreate the composition manually I get different results, see below.
**To Reproduce**
```python
from sktime.datasets import …