public interface Predictions
| Modifier and Type | Method and Description |
|---|---|
retrofit2.Call<ModelMetricsListSchemaV3> |
predict(java.lang.String model,
java.lang.String frame) |
retrofit2.Call<ModelMetricsListSchemaV3> |
predict(java.lang.String model,
java.lang.String frame,
java.lang.String predictions_frame,
java.lang.String deviances_frame,
boolean reconstruction_error,
boolean reconstruction_error_per_feature,
int deep_features_hidden_layer,
java.lang.String deep_features_hidden_layer_name,
boolean reconstruct_train,
boolean project_archetypes,
boolean reverse_transform,
boolean leaf_node_assignment,
ModelLeafNodeAssignmentLeafNodeAssignmentType leaf_node_assignment_type,
boolean predict_staged_proba,
boolean predict_contributions,
boolean row_to_tree_assignment,
ModelContributionsContributionsOutputFormat predict_contributions_output_format,
int top_n,
int bottom_n,
boolean compare_abs,
boolean feature_frequencies,
int exemplar_index,
boolean deviances,
java.lang.String custom_metric_func,
java.lang.String auc_type,
java.lang.String auuc_type,
double[] custom_auuc_thresholds,
int auuc_nbins,
java.lang.String background_frame,
boolean output_space,
boolean output_per_reference,
java.lang.String _exclude_fields)
Score (generate predictions) for the specified Frame with the specified Model.
|
retrofit2.Call<JobV3> |
predictAsync(java.lang.String model,
java.lang.String frame) |
retrofit2.Call<JobV3> |
predictAsync(java.lang.String model,
java.lang.String frame,
java.lang.String predictions_frame,
java.lang.String deviances_frame,
boolean reconstruction_error,
boolean reconstruction_error_per_feature,
int deep_features_hidden_layer,
java.lang.String deep_features_hidden_layer_name,
boolean reconstruct_train,
boolean project_archetypes,
boolean reverse_transform,
boolean leaf_node_assignment,
ModelLeafNodeAssignmentLeafNodeAssignmentType leaf_node_assignment_type,
boolean predict_staged_proba,
boolean predict_contributions,
boolean row_to_tree_assignment,
ModelContributionsContributionsOutputFormat predict_contributions_output_format,
int top_n,
int bottom_n,
boolean compare_abs,
boolean feature_frequencies,
int exemplar_index,
boolean deviances,
java.lang.String custom_metric_func,
java.lang.String auc_type,
java.lang.String auuc_type,
double[] custom_auuc_thresholds,
int auuc_nbins,
java.lang.String background_frame,
boolean output_space,
boolean output_per_reference,
java.lang.String _exclude_fields)
Score (generate predictions) for the specified Frame with the specified Model.
|
@FormUrlEncoded
@POST(value="/3/Predictions/models/{model}/frames/{frame}")
retrofit2.Call<ModelMetricsListSchemaV3> predict(@Path(value="model")
java.lang.String model,
@Path(value="frame")
java.lang.String frame,
@Field(value="predictions_frame")
java.lang.String predictions_frame,
@Field(value="deviances_frame")
java.lang.String deviances_frame,
@Field(value="reconstruction_error")
boolean reconstruction_error,
@Field(value="reconstruction_error_per_feature")
boolean reconstruction_error_per_feature,
@Field(value="deep_features_hidden_layer")
int deep_features_hidden_layer,
@Field(value="deep_features_hidden_layer_name")
java.lang.String deep_features_hidden_layer_name,
@Field(value="reconstruct_train")
boolean reconstruct_train,
@Field(value="project_archetypes")
boolean project_archetypes,
@Field(value="reverse_transform")
boolean reverse_transform,
@Field(value="leaf_node_assignment")
boolean leaf_node_assignment,
@Field(value="leaf_node_assignment_type")
ModelLeafNodeAssignmentLeafNodeAssignmentType leaf_node_assignment_type,
@Field(value="predict_staged_proba")
boolean predict_staged_proba,
@Field(value="predict_contributions")
boolean predict_contributions,
@Field(value="row_to_tree_assignment")
boolean row_to_tree_assignment,
@Field(value="predict_contributions_output_format")
ModelContributionsContributionsOutputFormat predict_contributions_output_format,
@Field(value="top_n")
int top_n,
@Field(value="bottom_n")
int bottom_n,
@Field(value="compare_abs")
boolean compare_abs,
@Field(value="feature_frequencies")
boolean feature_frequencies,
@Field(value="exemplar_index")
int exemplar_index,
@Field(value="deviances")
boolean deviances,
@Field(value="custom_metric_func")
java.lang.String custom_metric_func,
@Field(value="auc_type")
java.lang.String auc_type,
@Field(value="auuc_type")
java.lang.String auuc_type,
@Field(value="custom_auuc_thresholds")
double[] custom_auuc_thresholds,
@Field(value="auuc_nbins")
int auuc_nbins,
@Field(value="background_frame")
java.lang.String background_frame,
@Field(value="output_space")
boolean output_space,
@Field(value="output_per_reference")
boolean output_per_reference,
@Field(value="_exclude_fields")
java.lang.String _exclude_fields)
model - Key of Model of interest (optional)frame - Key of Frame of interest (optional)predictions_frame - Key of predictions frame, if predictions are requested (optional)deviances_frame - Key for the frame containing per-observation deviances (optional)reconstruction_error - Compute reconstruction error (optional, only for Deep Learning AutoEncoder models)reconstruction_error_per_feature - Compute reconstruction error per feature (optional, only for Deep
Learning AutoEncoder models)deep_features_hidden_layer - Extract Deep Features for given hidden layer (optional, only for Deep Learning
models)deep_features_hidden_layer_name - Extract Deep Features for given hidden layer by name (optional, only for
Deep Water models)reconstruct_train - Reconstruct original training frame (optional, only for GLRM models)project_archetypes - Project GLRM archetypes back into original feature space (optional, only for GLRM
models)reverse_transform - Reverse transformation applied during training to model output (optional, only for GLRM
models)leaf_node_assignment - Return the leaf node assignment (optional, only for DRF/GBM models)leaf_node_assignment_type - Type of the leaf node assignment (optional, only for DRF/GBM models)predict_staged_proba - Predict the class probabilities at each stage (optional, only for GBM models)predict_contributions - Predict the feature contributions - Shapley values (optional, only for DRF, GBM and
XGBoost models)row_to_tree_assignment - Return which row is used in which tree (optional, only for GBM models)predict_contributions_output_format - Specify how to output feature contributions in XGBoost - XGBoost by
default outputs contributions for 1-hot encoded features, specifying a
Compact output format will produce a per-feature contributiontop_n - Only for predict_contributions function - sort Shapley values and return top_n highest (optional)bottom_n - Only for predict_contributions function - sort Shapley values and return bottom_n lowest
(optional)compare_abs - Only for predict_contributions function - sort absolute Shapley values (optional)feature_frequencies - Retrieve the feature frequencies on paths in trees in tree-based models (optional,
only for GBM, DRF and Isolation Forest)exemplar_index - Retrieve all members for a given exemplar (optional, only for Aggregator models)deviances - Compute the deviances per row (optional, only for classification or regression models)custom_metric_func - Reference to custom evaluation function, format: `language:keyName=funcName`auc_type - Set default multinomial AUC type. Must be one of: "AUTO", "NONE", "MACRO_OVR", "WEIGHTED_OVR",
"MACRO_OVO", "WEIGHTED_OVO". Default is "NONE" (optional, only for multinomial classification).auuc_type - Set default AUUC type for uplift binomial classification. Must be one of: "AUTO", "qini",
"lift", "gain". Default is "AUTO" (optional, only for uplift binomial classification).custom_auuc_thresholds - Custom AUUC thresholds (for uplift binomial classification).auuc_nbins - Set number of bins to calculate AUUC. Must be -1 or higher than 0. Default is -1 which means
1000 (optional, only for uplift binomial classification).background_frame - Specify background frame used as a reference for calculating SHAP.output_space - If true, transform contributions so that they sum up to the difference in the output space
(applicable iff contributions are in link space). Note that this transformation is an
approximation and the contributions won't be exact SHAP values.output_per_reference - If true, return contributions against each background sample (aka reference), i.e.
phi(feature, x, bg), otherwise return contributions averaged over the background
sample (phi(feature, x) = E_{bg} phi(feature, x, bg))_exclude_fields - Comma-separated list of JSON field paths to exclude from the result, used like:
"/3/Frames?_exclude_fields=frames/frame_id/URL,__meta"@FormUrlEncoded
@POST(value="/3/Predictions/models/{model}/frames/{frame}")
retrofit2.Call<ModelMetricsListSchemaV3> predict(@Path(value="model")
java.lang.String model,
@Path(value="frame")
java.lang.String frame)
@FormUrlEncoded
@POST(value="/4/Predictions/models/{model}/frames/{frame}")
retrofit2.Call<JobV3> predictAsync(@Path(value="model")
java.lang.String model,
@Path(value="frame")
java.lang.String frame,
@Field(value="predictions_frame")
java.lang.String predictions_frame,
@Field(value="deviances_frame")
java.lang.String deviances_frame,
@Field(value="reconstruction_error")
boolean reconstruction_error,
@Field(value="reconstruction_error_per_feature")
boolean reconstruction_error_per_feature,
@Field(value="deep_features_hidden_layer")
int deep_features_hidden_layer,
@Field(value="deep_features_hidden_layer_name")
java.lang.String deep_features_hidden_layer_name,
@Field(value="reconstruct_train")
boolean reconstruct_train,
@Field(value="project_archetypes")
boolean project_archetypes,
@Field(value="reverse_transform")
boolean reverse_transform,
@Field(value="leaf_node_assignment")
boolean leaf_node_assignment,
@Field(value="leaf_node_assignment_type")
ModelLeafNodeAssignmentLeafNodeAssignmentType leaf_node_assignment_type,
@Field(value="predict_staged_proba")
boolean predict_staged_proba,
@Field(value="predict_contributions")
boolean predict_contributions,
@Field(value="row_to_tree_assignment")
boolean row_to_tree_assignment,
@Field(value="predict_contributions_output_format")
ModelContributionsContributionsOutputFormat predict_contributions_output_format,
@Field(value="top_n")
int top_n,
@Field(value="bottom_n")
int bottom_n,
@Field(value="compare_abs")
boolean compare_abs,
@Field(value="feature_frequencies")
boolean feature_frequencies,
@Field(value="exemplar_index")
int exemplar_index,
@Field(value="deviances")
boolean deviances,
@Field(value="custom_metric_func")
java.lang.String custom_metric_func,
@Field(value="auc_type")
java.lang.String auc_type,
@Field(value="auuc_type")
java.lang.String auuc_type,
@Field(value="custom_auuc_thresholds")
double[] custom_auuc_thresholds,
@Field(value="auuc_nbins")
int auuc_nbins,
@Field(value="background_frame")
java.lang.String background_frame,
@Field(value="output_space")
boolean output_space,
@Field(value="output_per_reference")
boolean output_per_reference,
@Field(value="_exclude_fields")
java.lang.String _exclude_fields)
model - Key of Model of interest (optional)frame - Key of Frame of interest (optional)predictions_frame - Key of predictions frame, if predictions are requested (optional)deviances_frame - Key for the frame containing per-observation deviances (optional)reconstruction_error - Compute reconstruction error (optional, only for Deep Learning AutoEncoder models)reconstruction_error_per_feature - Compute reconstruction error per feature (optional, only for Deep
Learning AutoEncoder models)deep_features_hidden_layer - Extract Deep Features for given hidden layer (optional, only for Deep Learning
models)deep_features_hidden_layer_name - Extract Deep Features for given hidden layer by name (optional, only for
Deep Water models)reconstruct_train - Reconstruct original training frame (optional, only for GLRM models)project_archetypes - Project GLRM archetypes back into original feature space (optional, only for GLRM
models)reverse_transform - Reverse transformation applied during training to model output (optional, only for GLRM
models)leaf_node_assignment - Return the leaf node assignment (optional, only for DRF/GBM models)leaf_node_assignment_type - Type of the leaf node assignment (optional, only for DRF/GBM models)predict_staged_proba - Predict the class probabilities at each stage (optional, only for GBM models)predict_contributions - Predict the feature contributions - Shapley values (optional, only for DRF, GBM and
XGBoost models)row_to_tree_assignment - Return which row is used in which tree (optional, only for GBM models)predict_contributions_output_format - Specify how to output feature contributions in XGBoost - XGBoost by
default outputs contributions for 1-hot encoded features, specifying a
Compact output format will produce a per-feature contributiontop_n - Only for predict_contributions function - sort Shapley values and return top_n highest (optional)bottom_n - Only for predict_contributions function - sort Shapley values and return bottom_n lowest
(optional)compare_abs - Only for predict_contributions function - sort absolute Shapley values (optional)feature_frequencies - Retrieve the feature frequencies on paths in trees in tree-based models (optional,
only for GBM, DRF and Isolation Forest)exemplar_index - Retrieve all members for a given exemplar (optional, only for Aggregator models)deviances - Compute the deviances per row (optional, only for classification or regression models)custom_metric_func - Reference to custom evaluation function, format: `language:keyName=funcName`auc_type - Set default multinomial AUC type. Must be one of: "AUTO", "NONE", "MACRO_OVR", "WEIGHTED_OVR",
"MACRO_OVO", "WEIGHTED_OVO". Default is "NONE" (optional, only for multinomial classification).auuc_type - Set default AUUC type for uplift binomial classification. Must be one of: "AUTO", "qini",
"lift", "gain". Default is "AUTO" (optional, only for uplift binomial classification).custom_auuc_thresholds - Custom AUUC thresholds (for uplift binomial classification).auuc_nbins - Set number of bins to calculate AUUC. Must be -1 or higher than 0. Default is -1 which means
1000 (optional, only for uplift binomial classification).background_frame - Specify background frame used as a reference for calculating SHAP.output_space - If true, transform contributions so that they sum up to the difference in the output space
(applicable iff contributions are in link space). Note that this transformation is an
approximation and the contributions won't be exact SHAP values.output_per_reference - If true, return contributions against each background sample (aka reference), i.e.
phi(feature, x, bg), otherwise return contributions averaged over the background
sample (phi(feature, x) = E_{bg} phi(feature, x, bg))_exclude_fields - Comma-separated list of JSON field paths to exclude from the result, used like:
"/3/Frames?_exclude_fields=frames/frame_id/URL,__meta"@FormUrlEncoded
@POST(value="/4/Predictions/models/{model}/frames/{frame}")
retrofit2.Call<JobV3> predictAsync(@Path(value="model")
java.lang.String model,
@Path(value="frame")
java.lang.String frame)