secretflow.ml.boost.sgb_v.factory.booster#

Classes:

GlobalOrdermapBooster(heu, tree_trainer)

This class provides both classification and regression tree boosting (also known as GBDT, GBM) for vertical split dataset setting by using level wise boost.

class secretflow.ml.boost.sgb_v.factory.booster.GlobalOrdermapBooster(heu: HEU, tree_trainer: TreeTrainer)[source]#

Bases: Composite

This class provides both classification and regression tree boosting (also known as GBDT, GBM) for vertical split dataset setting by using level wise boost.

Methods:

`__init__`(heu, tree_trainer)
`show_params`()
`set_params`(params)
`get_params`([params])
`set_devices`(devices)
`fit`(dataset, label)

__init__(heu: HEU, tree_trainer: TreeTrainer) → None[source]#

show_params()[source]#

set_params(params: dict)[source]#

get_params(params: dict = {}) → dict[source]#

set_devices(devices: Devices)[source]#

fit(dataset: Union[FedNdarray, VDataFrame], label: Union[FedNdarray, VDataFrame]) → SgbModel[source]#

secretflow.ml.boost.sgb_v.factory.booster.global_ordermap_booster#

Classes:

`GlobalOrdermapBoosterComponents`(...)
`GlobalOrdermapBoosterParams`([num_boost_round])	params specifically belonged to global ordermap booster, not its components.
`GlobalOrdermapBooster`(heu, tree_trainer)	This class provides both classification and regression tree boosting (also known as GBDT, GBM) for vertical split dataset setting by using level wise boost.

class secretflow.ml.boost.sgb_v.factory.booster.global_ordermap_booster.GlobalOrdermapBoosterComponents(preprocessor: secretflow.ml.boost.sgb_v.factory.components.data_preprocessor.data_preprocessor.DataPreprocessor = <secretflow.ml.boost.sgb_v.factory.components.data_preprocessor.data_preprocessor.DataPreprocessor object at 0x7f47491c7c70>, order_map_manager: secretflow.ml.boost.sgb_v.factory.components.order_map_manager.order_map_manager.OrderMapManager = <secretflow.ml.boost.sgb_v.factory.components.order_map_manager.order_map_manager.OrderMapManager object at 0x7f47491d6f40>, model_builder: secretflow.ml.boost.sgb_v.factory.components.model_builder.model_builder.ModelBuilder = <secretflow.ml.boost.sgb_v.factory.components.model_builder.model_builder.ModelBuilder object at 0x7f47491bcaf0>)[source]#

Bases: object

Attributes:

`preprocessor`
`order_map_manager`
`model_builder`

Methods:

__init__([preprocessor, order_map_manager, ...])

preprocessor: DataPreprocessor = <secretflow.ml.boost.sgb_v.factory.components.data_preprocessor.data_preprocessor.DataPreprocessor object>#

order_map_manager: OrderMapManager = <secretflow.ml.boost.sgb_v.factory.components.order_map_manager.order_map_manager.OrderMapManager object>#

model_builder: ModelBuilder = <secretflow.ml.boost.sgb_v.factory.components.model_builder.model_builder.ModelBuilder object>#

__init__(preprocessor: ~secretflow.ml.boost.sgb_v.factory.components.data_preprocessor.data_preprocessor.DataPreprocessor = <secretflow.ml.boost.sgb_v.factory.components.data_preprocessor.data_preprocessor.DataPreprocessor object>, order_map_manager: ~secretflow.ml.boost.sgb_v.factory.components.order_map_manager.order_map_manager.OrderMapManager = <secretflow.ml.boost.sgb_v.factory.components.order_map_manager.order_map_manager.OrderMapManager object>, model_builder: ~secretflow.ml.boost.sgb_v.factory.components.model_builder.model_builder.ModelBuilder = <secretflow.ml.boost.sgb_v.factory.components.model_builder.model_builder.ModelBuilder object>) → None#

class secretflow.ml.boost.sgb_v.factory.booster.global_ordermap_booster.GlobalOrdermapBoosterParams(num_boost_round: int = 10)[source]#

Bases: object

params specifically belonged to global ordermap booster, not its components.

num_boost_roundint, default=10: Number of boosting iterations. Same as number of trees. range: [1, 1024]

Attributes:

num_boost_round

Methods:

__init__([num_boost_round])

num_boost_round: int = 10#

__init__(num_boost_round: int = 10) → None#

class secretflow.ml.boost.sgb_v.factory.booster.global_ordermap_booster.GlobalOrdermapBooster(heu: HEU, tree_trainer: TreeTrainer)[source]#

Bases: Composite

This class provides both classification and regression tree boosting (also known as GBDT, GBM) for vertical split dataset setting by using level wise boost.

Methods:

`__init__`(heu, tree_trainer)
`show_params`()
`set_params`(params)
`get_params`([params])
`set_devices`(devices)
`fit`(dataset, label)

__init__(heu: HEU, tree_trainer: TreeTrainer) → None[source]#

show_params()[source]#

set_params(params: dict)[source]#

get_params(params: dict = {}) → dict[source]#

set_devices(devices: Devices)[source]#

fit(dataset: Union[FedNdarray, VDataFrame], label: Union[FedNdarray, VDataFrame]) → SgbModel[source]#