You are here:
Empty Predictions
Let’s say you build a prediction and get no prediction scores. Explore a possible reason.
You predict a number field that defaults to 0, and you use all records as examples.
When you select Use all records that have a value on a numeric prediction (Beta), Einstein provides scores only for records that don’t have a value for the predicted field.
If the predicted field has a default value of 0, no records remain to be scored. Edit your prediction to use specific records as examples instead, and set your condition logic to include only records that have a value not equal to 0 for the predicted field. Build your prediction again, and scores usually are available within 24 hours.
| If the numeric predicted field defaults to... | And your examples are set to... | Your prediction results are... |
|---|---|---|
| Null | Default (Use all records with a value for the predicted field) | Populated (Einstein provides scores on records) |
| 0 | Default (Use all records with a value for the predicted field) | Empty (Null; Einstein has no records left to score) |
| 0 | Only include records that don’t have a 0 value for the predicted field | Populated (Einstein provides scores on records) |
