Output Table Schema
Each row of the table represents one observation.
| Column | Data Type | Description |
|---|---|---|
| id | INTEGER | Row (observation) identifier. |
| prediction | VARCHAR | Prediction for observation. |
| loglik | DOUBLE PRECISION | [Appears only without Responses syntax element.] Loglikelihood that observation belongs to class label in prediction column. |
| prob | DOUBLE PRECISION | [Appears only when you both specify OutputProb ('true') and omit Responses.] Probability that observation belongs to class label in prediction column, which is max(softmax(loglik)). |
| loglik_response | DOUBLE PRECISION | [Appears only with Responses syntax element. Column appears once for each specified response.] Loglikelihood (natural logarithm of probability) that observation belongs to class response. |
| prob_response | DOUBLE PRECISION | [Appears only when you specify both OutputProb ('true') and Responses. Column appears once for each specified response.] Probability that observation belongs to class response, which is calculated as follows: softmax(loglik_response) |