secretflow.ml.boost.ss_xgb_v.core package

Submodules

secretflow.ml.boost.ss_xgb_v.core.node_split module

Classes:

RegType(value)

An enumeration.

Functions:

sigmoid(pred)
compute_obj(G, H, reg_lambda)	compute objective values of input buckets.
compute_weight(G, H, reg_lambda, learning_rate)	compute weight values of tree leaf nodes.
get_weight(context, s)	compute weight values of tree leaf nodes.
compute_gh(y, pred, objective)	compute first and second order gradient of each sample.
global_setup(buckets_map, y, seed, ...)	Set up global context.
tree_setup(context, pred, col_choices, ...)	Set up pre-tree context.
find_best_split_bucket(context, nodes_s, ...)	compute the gradient sums of the containing instances in each split bucket and find best split bucket for each node which has the max split gain.
init_pred(base, samples)
root_select(samples)
get_child_select(nodes_s, lchilds_ss)	compute the next level's select indexes.
predict_tree_weight(selects, weights)	get final pred for this tree.
do_leaf(context, ss)

class secretflow.ml.boost.ss_xgb_v.core.node_split.RegType(value)

Bases: Enum

An enumeration.

Attributes:

Linear
Logistic

Linear = 'linear'

Logistic = 'logistic'

secretflow.ml.boost.ss_xgb_v.core.node_split.sigmoid(pred: ndarray) → ndarray

secretflow.ml.boost.ss_xgb_v.core.node_split.compute_obj(G: ndarray, H: ndarray, reg_lambda: float) → ndarray

compute objective values of input buckets.

Parameters

• G/H – sum of first and second order gradient in each bucket.

• reg_lambda – L2 regularization term

Returns

objective values.

secretflow.ml.boost.ss_xgb_v.core.node_split.compute_weight(G: float, H: float, reg_lambda: float, learning_rate: float) → ndarray

compute weight values of tree leaf nodes.

Parameters

• G/H – sum of first and second order gradient in each node.

• reg_lambda – L2 regularization term

• learning_rate – Step size shrinkage used in update to prevents overfitting.

Returns

weight values.

secretflow.ml.boost.ss_xgb_v.core.node_split.get_weight(context: Dict[str, Any], s: ndarray) → ndarray

compute weight values of tree leaf nodes.

Parameters

• context – comparison context.

• s – sample selects in each leaf node.

Returns

weight values.

secretflow.ml.boost.ss_xgb_v.core.node_split.compute_gh(y: ndarray, pred: ndarray, objective: RegType) → Tuple[ndarray, ndarray]

compute first and second order gradient of each sample.

Parameters

• y – sample true label of each sample.

• pred – prediction of each sample.

• objective – regression learning objective,

Returns

weight values.

secretflow.ml.boost.ss_xgb_v.core.node_split.global_setup(buckets_map: List[ndarray], y: ndarray, seed: int, reg_lambda: float, learning_rate: float) → Dict[str, Any]

Set up global context.

secretflow.ml.boost.ss_xgb_v.core.node_split.tree_setup(context: Dict[str, Any], pred: ndarray, col_choices: List[ndarray], objective: RegType, samples: int, subsample: float) → Dict[str, Any]

Set up pre-tree context.

secretflow.ml.boost.ss_xgb_v.core.node_split.find_best_split_bucket(context: Dict[str, Any], nodes_s: List[ndarray], last_level: bool) → Tuple[ndarray, Dict[str, Any]]

compute the gradient sums of the containing instances in each split bucket and find best split bucket for each node which has the max split gain.

Parameters

• context – comparison context.

• nodes_s – sample select indexes of each node from same tree level.

• last_level – if this split is last level, next level is leaf nodes.

Returns

idx of split bucket for each node.

secretflow.ml.boost.ss_xgb_v.core.node_split.init_pred(base: float, samples: int)

secretflow.ml.boost.ss_xgb_v.core.node_split.root_select(samples: int) → List[ndarray]

secretflow.ml.boost.ss_xgb_v.core.node_split.get_child_select(nodes_s: List[ndarray], lchilds_ss: List[ndarray]) → List[ndarray]

compute the next level’s select indexes.

Parameters

• nodes_s – sample select indexes of each node from current level’s nodes.

• lchilds_ss – left children’s sample select idx for current level’s nodes.

Returns

sample select indexes for nodes in next tree level.

secretflow.ml.boost.ss_xgb_v.core.node_split.predict_tree_weight(selects: List[ndarray], weights: ndarray) → ndarray

get final pred for this tree.

Parameters

• selects – leaf nodes’ sample selects from each model handler.

• weights – leaf weights in secure share.

Returns

pred

secretflow.ml.boost.ss_xgb_v.core.node_split.do_leaf(context: Dict[str, Any], ss: List[ndarray]) → ndarray

secretflow.ml.boost.ss_xgb_v.core.tree_worker module

Classes:

XgbTreeWorker

alias of ActorProxy(XgbTreeWorker)

secretflow.ml.boost.ss_xgb_v.core.tree_worker.XgbTreeWorker

alias of ActorProxy(XgbTreeWorker) Methods:

__init__(*args, **kwargs)	Abstraction device object base class.
predict_weight_select(x, tree, *[, ...])	computer leaf nodes' sample selects known by this partition.
build_maps(x, *[, _ray_trace_ctx])	split features into buckets and build maps use in train.
global_setup(x, buckets, seed, *[, ...])	Set up global context.
update_buckets_count(buckets_count, *[, ...])	save how many buckets in each partition's all features.
tree_setup(colsample, *[, _ray_trace_ctx])	Set up tree context and do col sample if colsample < 1
tree_finish(*[, _ray_trace_ctx])
do_split(split_buckets, *[, _ray_trace_ctx])	record split info and generate next level's left children select.

secretflow.ml.boost.ss_xgb_v.core.utils module

Functions:

prepare_dataset(ds)

check data setting and get total shape.

secretflow.ml.boost.ss_xgb_v.core.utils.prepare_dataset(ds: Union[FedNdarray, VDataFrame]) → Tuple[FedNdarray, Tuple[int, int]]

check data setting and get total shape.

Parameters

ds – input dataset

Returns

dataset in unified type Second: shape concat all partition.

Return type

First

secretflow.ml.boost.ss_xgb_v.core.xgb_tree module

Classes:

XgbTree()

class secretflow.ml.boost.ss_xgb_v.core.xgb_tree.XgbTree

Bases: object

Methods:

__init__()
insert_split_node(feature, value)

__init__() → None

insert_split_node(feature: int, value: float) → None

Module contents