推断阶段。
-
场景六:当模型名在系统表中查找不到,数据库会报ERROR:
openGauss=# select id, PREDICT BY patient_logistic_regression (FEATURES second_attack,treatment) FROM patients; ERROR: There is no model called "patient_logistic_regression".
-
场景七:当做推断任务FEATURES的数据维度和数据类型与训练集存在不一致,将报ERROR,并提示错误原因,例如:
openGauss=# select id, PREDICT BY patient_linear_regression (FEATURES second_attack) FROM patients; ERROR: Invalid number of features for prediction, provided 1, expected 2 CONTEXT: referenced column: patient_linear_regression_pred ------------------------------------------------------------------------------------------------------------------------------------- openGauss=# select id, PREDICT BY patient_linear_regression (FEATURES 1,second_attack,treatment) FROM patients; ERROR: Invalid number of features for prediction, provided 3, expected 2 CONTEXT: referenced column: patient_linear_regression_pre
详情查看:https://opengauss.org
详情查看:https://docs-opengauss.osinfra.cn