By default, the KNN function outputs a table to the screen. If you specify an output table name, the function outputs a message to the screen and creates an output table with the specified name.
Output Message Schema
Column | Data Type | Description |
---|---|---|
message | VARCHAR | Reports that final result is stored in table specified in OutputTable syntax element. |
OutputTable Schema, OutputProb ('false') (Default)
Column | Data Type | Description |
---|---|---|
test_id_column | Same as in TestData table | Unique test data object identifier. |
prediction | VARCHAR | Prediction from TrainingData response_column to which function mapped test data object. |
accumulate_column | Same as in TestData table | Column copied from TestData table. |
OutputTable Schema, OutputProb ('true') and Responses Omitted
Column | Data Type | Description |
---|---|---|
test_id_column | Same as in TestData table | Unique test data object identifier. |
prediction | VARCHAR | Prediction from TrainingData response_column to which function mapped test data object. |
prob | DOUBLE PRECISION | Probability that observation belongs to predicted category, which is based on weighted voting (categories) of k nearest neighbors. |
accumulate_column | Same as in TestData table | Column copied from TestData table. |
OutputTable Schema, OutputProb ('true') and Responses Specified
Column | Data Type | Description |
---|---|---|
test_id_column | Same as in TestData table | Unique test data object identifier. |
prediction | VARCHAR | Category from TrainingData response_column to which function mapped test data object. |
prob_response | DOUBLE PRECISION | [Column appears once for each specified response.] Probability that observation belongs to belongs to category response, which is based on weighted voting (categories) of k nearest neighbors. |
accumulate_column | Same as in TestData table | Column copied from TestData table. |