Dynamic Prediction Dataset Naming

AutoML lets you add time-based variables to your synced prediction dataset names.

What is dynamic prediction dataset naming?
Dynamic prediction dataset naming is available for all predictions EXCEPT those powered by CSV file upload. (.CSV files coming from buckets like Amazon S3 are supported.) When you create a prediction dataset and enable Prediction Sync, you can specify the name you'd like for the predicted dataset and include a variety of time-based variables that AutoML will replace with actual values when your prediction dataset is synced. Depending on the variables that you use, this capability allows you to keep a running history of your predictions in your data provider and prevent AutoML from overwriting your prediction dataset every time it is synced to your data provider.

How does it work?
Dynamic prediction dataset naming uses time-based variables, including:

• {{Year}}
• {{Month}}
• {{Day}}
• {{Hour}}
• {{Minute}}
• {{Second}}

When you create a prediction dataset and enable Prediction Sync, simply insert one or more variables into the prediction dataset name.

For some examples, let's assume that you're creating and enabling a synced prediction dataset on August 24, 2021 at 5:31:15pm...

Important considerations about dynamic prediction dataset naming

• This one's REALLY important: Before using dynamic prediction dataset naming, you should fully read this article to ensure you understand how Prediction Sync works.
• Double curly braces must be included around each of the variables in your prediction dataset name, or AutoML will not replace the variables with actual values.
• When you first enable a synced prediction using dynamic prediction dataset naming, AutoML will immediately write a prediction dataset to your data provider using the variable values at the time you enable prediction sync. After that, the dataset will be created (or overwritten, depending on the variables you used) whenever your predictions are scheduled to sync (by default, AutoML users have a daily prediction schedule).
• AutoML will overwrite your prediction dataset if the variable(s) used don't change from one prediction sync to the next. For example, if you put {{Year}} as the only variable in the prediction dataset name, AutoML will overwrite the prediction dataset with every sync through the end of the calendar year.
• Since AutoML will NOT overwrite a dataset in your data provider that exists before you enable Prediction Sync, if you select variables that result in a prediction dataset name that happens to already exist in your provider before you enable Prediction Sync, the prediction sync will fail. For example, if it's 2021 and if you already have a table in your database called SALES_2021 and you create a prediction dataset name of SALES_{{Year}} when you enable Prediction Sync it will fail, since the resulting dataset name of SALES_2021 existed in your data provider before you enabled Prediction Sync.