Closed jpzhangvincent closed 4 years ago
Oh I realized the testing time series data frame is a tsibble
object. I have to convert to tibble
object and then it works.
Yes, you'll need to work exclusively with data.frame
or tibble
objects. tsibble
have different properties making it difficult to predict how they will respond to the modeltime workflow.
I installed the development version from GitHub and tried to forecast a time series using 'auto.rima' for intermittent demand forecasting. But I got the following error.
Here is my testing example. test_ts.csv.zip