Given predicted values (target for regression, class-1 probabilities or binomial or per-class probabilities for multinomial), compute a model metrics object

h2o.make_metrics( predicted, actuals, domain = NULL, distribution = NULL, weights = NULL, treatment = NULL, auc_type = "NONE", auuc_type = "AUTO", auuc_nbins = -1, custom_auuc_thresholds = NULL )

predicted | An H2OFrame containing predictions |
---|---|

actuals | An H2OFrame containing actual values |

domain | Vector with response factors for classification. |

distribution | Distribution for regression. |

weights | (optional) An H2OFrame containing observation weights. |

treatment | (optional, for uplift models only) An H2OFrame containing treatment column for uplift classification. |

auc_type | (optional) For multinomial classification you have to specify which type of agregated AUC/AUCPR will be used to calculate this metric. |

auuc_type | (optional) For uplift binomial classification you have to specify which type of AUUC will be used to calculate this metric. Possibilities are gini, lift, gain, AUTO. Default is AUTO which means qini. |

auuc_nbins | (optional) For uplift binomial classification you can specify number of bins to be used for calculation the AUUC. Default is -1, which means 1000. |

custom_auuc_thresholds | (optional) For uplift binomial classification you can specify exact thresholds to calculate AUUC. Default is NULL. If the thresholds are not defined, auuc_nbins will be used to calculate new thresholds from the predicted data. |

Returns an object of the H2OModelMetrics subclass.

if (FALSE) { library(h2o) h2o.init() prostate_path <- system.file("extdata", "prostate.csv", package = "h2o") prostate <- h2o.uploadFile(path = prostate_path) prostate$CAPSULE <- as.factor(prostate$CAPSULE) prostate_gbm <- h2o.gbm(3:9, "CAPSULE", prostate) pred <- h2o.predict(prostate_gbm, prostate)[, 3] ## class-1 probability h2o.make_metrics(pred, prostate$CAPSULE) }