Calculates one model performance metric on a dataset, comparing a prediction field with an actual field. Supported metrics are accuracy, precision, recall, F1 score, AUC-ROC, log loss, MAE, MSE, RMSE, and MAPE.

When to use it

Use it to measure how well a model performs on the data in a dataset. Add dimensions to get one result per group.

Required parameters

• Request body: a metric payload. Set metric to one of ACCURACY, PRECISION, RECALL, F1_SCORE, AUC_ROC, LOG_LOSS, MAE, MSE, RMSE, or MAPE, and give the dataset_id, the prediction_field, the actual_field, and optional filters and dimensions.

Response

A successful call returns 200 OK with metric and results, a list of values, each with its dimensions_values.

Related: Calculate several performance metrics, Compare two distributions, Run a SQL query on datasets