merlin.systems.dag.ops.fil.PredictForest

merlin.systems.dag.ops.fil.PredictForest#

class merlin.systems.dag.ops.fil.PredictForest(model, input_schema, *, backend='python', **fil_params)[source]#

Bases: InferenceOperator

Operator for running inference on Forest models.

This works for gradient-boosted decision trees (GBDTs) and Random forests (RF). While RF and GBDT algorithms differ in the way they train the models, they both produce a decision forest as their output.

Uses the Forest Inference Library (FIL) backend for inference.

__init__(model, input_schema, *, backend='python', **fil_params)[source]#

Instantiate a FIL inference operator.

Parameters:

  • model (Forest Model Instance) – A forest model class. Supports XGBoost, LightGBM, and Scikit-Learn.

  • input_schema (merlin.schema.Schema) – The schema representing the input columns expected by the model.

  • backend (str) – The Triton backend to use to when running this operator.

  • **fil_params – The parameters to pass to the FIL operator.

Methods

__init__(model, input_schema, *[, backend])

Instantiate a FIL inference operator.

column_mapping(col_selector)

Compute which output columns depend on which input columns

compute_column_schema(col_name, input_schema)

compute_input_schema(root_schema, ...)

Return the input schema representing the input columns this operator expects to use.

compute_output_schema(input_schema, col_selector)

Return the output schema representing the columns this operator returns.

compute_selector(input_schema, selector[, ...])

Provides a hook method for sub-classes to override to implement custom column selection logic.

create_node(selector)

_summary_

export(path, input_schema, output_schema[, ...])

Export the class object as a config and all related files to the user defined path.

from_model_registry(registry, **kwargs)

Loads the InferenceOperator from the provided ModelRegistry.

from_path(path, **kwargs)

Loads the InferenceOperator from the path where it was exported after training.

load_artifacts(artifact_path)

Load artifacts from disk required for operator function.

output_column_names(col_selector)

Given a set of columns names returns the names of the transformed columns this operator will produce

save_artifacts(artifact_path)

Save artifacts required to be reload operator state from disk

transform(col_selector, transformable)

Transform the dataframe by applying this FIL operator to the set of input columns.

validate_schemas(parents_schema, ...[, ...])

Provides a hook method that sub-classes can override to implement schema validation logic.

Attributes

dependencies

Defines an optional list of column dependencies for this operator.

dynamic_dtypes

export_name

Provides a clear common english identifier for this operator.

is_subgraph

label

output_dtype

output_properties

output_tags

scalar_shape

supported_formats

supports

Returns what kind of data representation this operator supports
compute_output_schema(input_schema: Schema, col_selector: ColumnSelector, prev_output_schema: Optional[Schema] = None) → Schema[source]#

Return the output schema representing the columns this operator returns.

compute_input_schema(root_schema: Schema, parents_schema: Schema, deps_schema: Schema, selector: ColumnSelector) → Schema[source]#

Return the input schema representing the input columns this operator expects to use.

transform(col_selector: ColumnSelector, transformable: Transformable) → Transformable[source]#

Transform the dataframe by applying this FIL operator to the set of input columns.

Parameters:

df (TensorTable) – A pandas or cudf dataframe that this operator will work on

Returns:

Returns a transformed dataframe for this operator

Return type:

TensorTable

Contents