hummingbird.ml._parse

All functions used for parsing input models are listed here. Some code here have been copied from https://github.com/onnx/sklearn-onnx/.

hummingbird.ml._parse._build_sklearn_api_parsers_map()

hummingbird.ml._parse._build_sparkml_api_parsers_map()

hummingbird.ml._parse._declare_input_variables(topology, raw_model_container, extra_config)

hummingbird.ml._parse._declare_output_variables(raw_model_container, extra_config, outputs)

hummingbird.ml._parse._fetch_input_slice(topology, inputs, column_indices)

Taken from https://github.com/onnx/sklearn-onnx/blob/9939c089a467676f4ffe9f3cb91098c4841f89d8/skl2onnx/_parse.py#L53.

hummingbird.ml._parse._get_column_index(i, inputs)

Taken from https://github.com/onnx/sklearn-onnx/blob/9939c089a467676f4ffe9f3cb91098c4841f89d8/skl2onnx/common/utils.py#L50. Returns a tuples (variable index, column index in that variable). The function has two different behaviours, one when i (column index) is an integer, another one when i is a string (column name). If i is a string, the function looks for input name with this name and returns (index, 0). If i is an integer, let’s assume first we have two inputs I0 = FloatTensorType([None, 2]) and I1 = FloatTensorType([None, 3]), in this case, here are the results:

get_column_index(0, inputs) -> (0, 0)
get_column_index(1, inputs) -> (0, 1)
get_column_index(2, inputs) -> (1, 0)
get_column_index(3, inputs) -> (1, 1)
get_column_index(4, inputs) -> (1, 2)

hummingbird.ml._parse._get_column_indices(indices, inputs, multiple=False)

Taken from https://github.com/onnx/sklearn-onnx/blob/9939c089a467676f4ffe9f3cb91098c4841f89d8/skl2onnx/common/utils.py#L105. Returns the requested graph inpudes based on their indices or names. See _parse._get_column_index. Args:

indices: variables indices or names inputs: model inputs

Returns:

a tuple (variable name, list of requested indices) if multiple is False, a dictionary { var_index: [ list of requested indices ] } if multiple is True

hummingbird.ml._parse._parse_onnx_api(topology, model, inputs)

This function handles all input ONNX models.

Args:

topology: The onnxconverter_common.topology.Topology where the model will be added model: A ONNX model object inputs: A list of `onnxconverter_common.topology.Variable`s

Returns:

A list of output onnxconverter_common.topology.Variable which will be passed to next stage

hummingbird.ml._parse._parse_onnx_single_operator(topology, operator)

This function handles the parsing of all ONNX operators.

Args:

topology: The onnxconverter_common.topology.Topology where the model will be added model: An ONNX operator

hummingbird.ml._parse._parse_sklearn_api(topology, model, inputs)

This is a delegate function adding the model to the input topology. It does nothing but invokes the correct parsing function according to the input model’s type.

Args:

topology: The hummingbitd.ml._topology.Topology object where the model will be added model: A scikit-learn model object inputs: A list of `onnxconverter_common.topology.Variable`s

Returns:

The output `onnxconverter_common.topology.Variable`s produced by the input model

hummingbird.ml._parse._parse_sklearn_bagging(topology, model, inputs)

Taken from https://github.com/onnx/sklearn-onnx/blob/9939c089a467676f4ffe9f3cb91098c4841f89d8/skl2onnx/_parse.py#L238. :param topology: Topology object :param model: A scikit-learn BaggingClassifier or BaggingRegressor object :param inputs: A list of Variable objects :return: A list of output variables produced by column transformer

hummingbird.ml._parse._parse_sklearn_column_transformer(topology, model, inputs)

Taken from https://github.com/onnx/sklearn-onnx/blob/9939c089a467676f4ffe9f3cb91098c4841f89d8/skl2onnx/_parse.py#L238. :param topology: Topology object :param model: A scikit-learn ColumnTransformer object :param inputs: A list of Variable objects :return: A list of output variables produced by column transformer

hummingbird.ml._parse._parse_sklearn_feature_union(topology, model, inputs)

Taken from https://github.com/onnx/sklearn-onnx/blob/9939c089a467676f4ffe9f3cb91098c4841f89d8/skl2onnx/_parse.py#L199. :param topology: Topology object :param model: A scikit-learn FeatureUnion object :param inputs: A list of Variable objects :return: A list of output variables produced by feature union

hummingbird.ml._parse._parse_sklearn_model_selection(topology, model, inputs)

Parameters:

• topology – Topology object

• model – A sklearn.model_selection object

• inputs – A list of Variable objects

Returns:

Output produced by sklearn.model_selection.* object

hummingbird.ml._parse._parse_sklearn_multi_output_regressor(topology, model, inputs)

Parameters:

• topology – Topology object

• model – A scikit-learn MultiOutputRegressor object

• inputs – A list of Variable objects

Returns:

Output produced by MultiOutputRegressor

hummingbird.ml._parse._parse_sklearn_pipeline(topology, model, inputs)

The basic ideas of scikit-learn parsing:

1. Sequentially go though all stages defined in the considered scikit-learn pipeline

2. The output variables of one stage will be fed into its next stage as the inputs.

Parameters:

• topology – Topology object defined in _topology.py

• model – scikit-learn pipeline object

• inputs – A list of Variable objects

Returns:

A list of output variables produced by the input pipeline

hummingbird.ml._parse._parse_sklearn_regressor_chain(topology, model, inputs)

Parameters:

• topology – Topology object

• model – A scikit-learn RegressorChain object

• inputs – A list of Variable objects

Returns:

Output produced by RegressorChain

hummingbird.ml._parse._parse_sklearn_single_model(topology, model, inputs)

This function handles all sklearn objects composed by a single model.

Args:

topology: The hummingbitd.ml._topology.Topology object where the model will be added model: A scikit-learn model object inputs: A list of `onnxconverter_common.topology.Variable`s

Returns:

A list of output onnxconverter_common.topology.Variable which will be passed to next stage

hummingbird.ml._parse._parse_sklearn_stacking(topology, model, inputs)

Taken from https://github.com/onnx/sklearn-onnx/blob/9939c089a467676f4ffe9f3cb91098c4841f89d8/skl2onnx/_parse.py#L238. :param topology: Topology object :param model: A scikit-learn Stacking object :param inputs: A list of Variable objects :return: A list of output variables produced by column transformer

hummingbird.ml._parse._parse_sparkml_api(topology, model, inputs)

This function handles all input Spark-ML models.

Args:

topology: The onnxconverter_common.topology.Topology where the model will be added model: A Spark-ML model object inputs: A list of `onnxconverter_common.topology.Variable`s

Returns:

A list of output onnxconverter_common.topology.Variable which will be passed to next stage

hummingbird.ml._parse._parse_sparkml_pipeline(topology, model, all_outputs)

The basic ideas of Spark-ML parsing:

1. Sequentially go though all stages defined in the considered Spark-ML pipeline passing all outputs that has been generated so far as the input. Operator will pick which inputs to operates on.

2. The output variables of one stage will be fed into its next stage as the inputs.

Parameters:

• topology – Topology object defined in _topology.py

• model – Spark-ML pipeline object

• all_outputs – A list of Variable objects

Returns:

A list of output variables produced by the input pipeline

hummingbird.ml._parse._parse_sparkml_single_operator(topology, operator, all_inputs)

This function handles the parsing of all Spark-ML operators.

Args:

topology: The onnxconverter_common.topology.Topology where the model will be added model: A Spark-ML operator all_inputs: A list of `onnxconverter_common.topology.Variable`s

hummingbird.ml._parse._remove_zipmap(node_list)

Method used to remove ZipMap operators in the graph.

hummingbird.ml._parse.parse_onnx_api_model(model)

Puts ONNX object into an abstract representation so that our framework can work seamlessly on models created with different machine learning tools.

Args:

model: A model object in onnx format

Returns:

A onnxconverter_common.topology.Topology object representing the input model

hummingbird.ml._parse.parse_sklearn_api_model(model, extra_config={})

Puts scikit-learn object into an abstract representation so that our framework can work seamlessly on models created with different machine learning tools.

Args:

model: A model object in scikit-learn format

Returns:

A onnxconverter_common.topology.Topology object representing the input model

hummingbird.ml._parse.parse_sparkml_api_model(model, extra_config={})

Puts Spark-ML object into an abstract representation so that our framework can work seamlessly on models created with different machine learning tools.

Args:

model: A model object in Spark-ML format

Returns:

A onnxconverter_common.topology.Topology object representing the input model