Scikit-Learn Exporter Module

def any_in(seq_a, seq_b)

Checks for common elements in two given sequence elements

Parameters:

• seq_a (list) – A list of items
• seq_b (list) – A list of items

Returns:

Return type:

Returns a boolean value if any item of seq_a belongs to seq_b or visa versa

def avgPathLength(n)

Generates average path length for Isolation forest models

Parameters:n (int) – Number of samples Returns: Return type:The average path length

def get_PMML_kwargs(model, derived_col_names, col_names, target_name, mining_imp_val, categoric_values, model_name)

It returns all the pmml elements.

Parameters:

• model (Scikit-learn model object) – An instance of Scikit-learn model.
• derived_col_names (List) – Contains column names after preprocessing
• col_names (List) – Contains list of feature/column names.
• target_name (String) – Name of the target column .
• mining_imp_val (tuple) – Contains the mining_attributes,mining_strategy, mining_impute_value
• categoric_values (tuple) – Contains Categorical attribute names and its values
• model_name (string) – Name of the model

Returns:

algo_kwargs – Get the PMML model argument based on scikit learn model object

Return type:

Dictionary

def get_anomaly_detection_output(model)

Generates output for anomaly detection models

Parameters:model – Scikit-learn’s model object Returns:Returns Nyoka’s Output object Return type:output_fields

def get_anomalydetection_model(model, derived_col_names, col_names, target_name, mining_imp_val, categoric_values, model_name)

Creates xml elements for anomaly detction models

Parameters:

• model – An instance of Scikit-learn model.
• derived_col_names (List) – Contains column names after preprocessing
• col_names (List) – Contains list of feature/column names.
• target_name (String) – Name of the Target column.
• mining_imp_val (tuple) – Contains the mining_attributes,mining_strategy, mining_impute_value
• categoric_values (tuple) – Contains Categorical attribute names and its values
• model_name (string) – Name of the model

Returns:

anomaly_detection_model – Returns Nyoka’s AnomalyDetectionModel object

Return type:

List

def get_bayes_inputs(model, derived_col_names)

It returns the Bayes Input element of the naive bayes model .

Parameters:

• model – An instance of Scikit-learn model.
• derived_col_names (List) – Contains column names after preprocessing.

Returns:

Returns Nyoka’s BayesInput object.

Return type:

bayes_inputs

def get_bayes_output(model, target_name)

It returns the Bayes Output element of the model

Parameters:

• model – An instance of Scikit-learn model.
• target_name (String) – Name of the Target column.

Returns:

Returns Nyoka’s BayesOutput object

Return type:

BayesOutput

def get_classificationMethod(model)

It returns the Classification method name for SVM models.

Parameters:model – A Scikit-learn model instance. Returns: Return type:Returns the classification method of the SVM model

def get_cluster_num(model)

Returns number of cluster for clustering models

Parameters:model – An instance of Scikit-learn model. Returns:model.n_clusters – Returns the number of clusters Return type:Integer

def get_cluster_vals(model, counts)

Generates cluster information for clustering models

Parameters:model – An instance of Scikit-learn model. Returns:cluster_flds – Returns Nyoka’s Cluster object Return type:List

def get_clustering_flds(col_names)

Generates cluster fields for clustering models

Parameters:col_names – Contains list of feature/column names. Returns:clustering_flds – Returns Nyoka’s ClusteringField object Return type:List

def get_clustering_model(model, derived_col_names, col_names, target_name, mining_imp_val, categoric_values, model_name)

Generates PMML elements for clustering models

Parameters:

• model – An instance of Scikit-learn model.
• derived_col_names (List) – Contains column names after preprocessing
• col_names (List) – Contains list of feature/column names.
• target_name (String) – Name of the Target column.
• mining_imp_val (tuple) – Contains the mining_attributes,mining_strategy, mining_impute_value
• categoric_values (tuple) – Contains Categorical attribute names and its values
• model_name (string) – Name of the model

Returns:

clustering_models – Returns Nyoka’s ClusteringModel object

Return type:

List

def get_comp_measure()

Generates comparison measure information for clustering models

Returns: Return type:Returns Nyoka’s ComparisonMeasure object

def get_comparison_measure(model)

It return the Comparison measure element for nearest neighbour model.

Parameters:model – An instance of Scikit-learn model. Returns:Returns Nyoka’s ComparisonMeasure object. Return type:comp_measure

def get_data_dictionary(model, feature_names, target_name, categoric_values)

It returns the Data Dictionary element.

Parameters:

• model – A Scikit-learn model instance.
• feature_names (List) – Contains the list of feature/column name.
• target_name (List) – Name of the Target column.
• categoric_values (tuple) – Contains Categorical attribute names and its values

Returns:

Returns Nyoka’s DataDictionary object

Return type:

data_dict

def get_dtype(feat_value)

It return the data type of the value.

Parameters:feat_value – Contains a value for finding the its data type. Returns: Return type:Returns the respective data type of that value.

def get_ensemble_models(model, derived_col_names, col_names, target_name, mining_imp_val, categoric_values, model_name)

Generates PMML elemenets for ensemble models

Parameters:

• model – An instance of Scikit-learn model.
• derived_col_names (List) – Contains column names after preprocessing.
• col_names (List) – Contains list of feature/column names.
• target_name (String) – Name of the Target column.
• mining_imp_val (tuple) – Contains the mining_attributes,mining_strategy, mining_impute_value.
• categoric_values (tuple) – Contains Categorical attribute names and its values
• model_name (string) – Name of the model

Returns:

mining_models – Returns Nyoka’s MiningModel object

Return type:

List

def get_funct(sk_model)

It returns the activation fucntion for a neural network model.

Parameters:model – A Scikit-learn model instance. Returns:a_fn – Returns the activation function. Return type:String

def get_header(description)

It returns the Header element of the pmml.

header :

Returns Nyoka’s Header object.

def get_inline_table(model)

It Returns the Inline Table element of the model.

Parameters:model – An instance of Scikit-learn model. Returns:Returns Nyoka’s InlineTable object Return type:InlineTable

def get_inner_segments(model, derived_col_names, col_names, index)

It returns the segments of a Segmentation.

Parameters:

• model – A Scikit-learn model instance.
• derived_col_names (List) – Contains column names after preprocessing.
• col_names (List) – Contains list of feature/column names.
• index (Integer) – The index of the estimator for the model

Returns:

segments – Nyoka’s Segment object

Return type:

List

def get_instance_fields(derived_col_names, target_name)

It returns the Instance field element.

Parameters:

• derived_col_names (List) – Contains column names after preprocessing.
• target_name (String) – Name of the Target column.

Returns:

Returns Nyoka’s InstanceFields object

Return type:

InstanceFields

def get_kernel_type(model)

It returns the kernel type element.

Parameters:model – A Scikit-learn model instance. Returns:kernel_kwargs – Get the respective kernel type of the SVM model. Return type:Dictionary

def get_knn_inputs(col_names)

It returns the KNN Inputs element.

Parameters:col_names (List) – Contains list of feature/column names. Returns:Returns Nyoka’s KNNInputs object. Return type:KNNInputs

def get_mining_func(model)

It returns the name of the mining function of the model.

Parameters:model – A Scikit-learn model instance. Returns:func_name – Returns the function name of the model Return type:String

def get_mining_schema(model, feature_names, target_name, mining_imp_val, categoric_values)

It returns the Mining Schema of the model.

Parameters:

• model – A Scikit-learn model instance.
• feature_names (List) – Contains the list of feature/column name.
• target_name (String) – Name of the Target column.
• mining_imp_val (tuple) – Contains the mining_attributes,mining_strategy, mining_impute_value.
• categoric_values (tuple) – Contains Categorical attribute names and its values

Returns:

Nyoka’s MiningSchema object

Return type:

MiningSchema

def get_model_kwargs(model, col_names, target_name, mining_imp_val, categoric_values)

It returns all the model element for a specific model.

Parameters:

• model – An instance of Scikit-learn model.
• col_names (List) – Contains list of feature/column names.
• target_name (String) – Name of the Target column.
• mining_imp_val (tuple) – Contains the mining_attributes,mining_strategy, mining_impute_value
• categoric_values (tuple) – Contains Categorical attribute names and its values

Returns:

model_kwargs – Returns function name, MiningSchema and Output of the sk_model object

Return type:

Dictionary

def get_multiple_model_method(model)

It returns the type of multiple model method for MiningModels.

Parameters:model – A Scikit-learn model instance Returns: Return type:The multiple model method for a MiningModel.

def get_naiveBayesModel(model, derived_col_names, col_names, target_name, mining_imp_val, categoric_values, model_name)

Generates PMML elements for naive bayes models

Parameters:

• model – An instance of Scikit-learn model.
• derived_col_names (List) – Contains column names after preprocessing.
• col_names (List) – Contains list of feature/column names.
• target_name (String) – Name of the Target column.
• mining_imp_val (tuple) – Contains the mining_attributes,mining_strategy, mining_impute_value
• categoric_values (tuple) – Contains Categorical attribute names and its values
• model_name (string) – Name of the model

Returns:

naive_bayes_model – Returns Nyoka’s NaiveBayesModel

Return type:

List

def get_nearestNeighbour_model(model, derived_col_names, col_names, target_name, mining_imp_val, categoric_values, model_name)

Generates PMML elements for nearest neighbour model

Parameters:

• model – An instance of Scikit-learn model.
• derived_col_names (List) – Contains column names after preprocessing
• col_names (List) – Contains list of feature/column names.
• target_name (String) – Name of the Target column.
• mining_imp_val (tuple) – Contains the mining_attributes,mining_strategy, mining_impute_value
• categoric_values (tuple) – Contains Categorical attribute names and its values
• model_name (string) – Name of the model

Returns:

Returns Nyoka’s NearestNeighborModel object

Return type:

nearest_neighbour_model

def get_neural_layer(model, feature_names, target_name)

It returns the Neural Layer and Neural Ouptput element.

Parameters:

• model – A Scikit-learn model instance.
• feature_names (List) – Contains the list of feature/column name.
• target_name (String) – Name of the Target column.

Returns:

• all_neuron_layer (List) – Nyoka’s NeuralLayer object
• neural_output_element – Nyoka’s NeuralOutput object

def get_neural_models(model, derived_col_names, col_names, target_name, mining_imp_val, categoric_values, model_name)

Generates PMML elements for neural network models

Parameters:

• model – A Scikit-learn model instance.
• derived_col_names (List) – Contains column names after preprocessing.
• col_names (List) – Contains list of feature/column names.
• target_name (String) – Name of the Target column.
• mining_imp_val (tuple) – Contains the mining_attributes,mining_strategy, mining_impute_value.
• categoric_values (tuple) – Contains Categorical attribute names and its values
• model_name (string) – Name of the model

Returns:

neural_model – Nyoka’s NeuralNetwork object

Return type:

List

def get_neuron_input(feature_names)

It returns the Neural Input element.

Parameters:feature_names (List) – Contains the list of feature/column name. Returns:Returns Nyoka’s NeuralInput object Return type:neural_input_element

def get_node(model, features_names, main_model=None)

It return the Node element of the model.

Parameters:

• model – An instance of the estimator of the tree object.
• features_names (List) – Contains the list of feature/column name.
• main_model – A Scikit-learn model instance.

Returns:

Return type:

Get all the underlying Nodes.

def get_outer_segmentation(model, derived_col_names, col_names, target_name, mining_imp_val, categoric_values, model_name)

It returns the Segmentation element of a MiningModel.

Parameters:

• model – A Scikit-learn model instance.
• derived_col_names (List) – Contains column names after preprocessing.
• col_names (List) – Contains list of feature/column names.
• target_name (String) – Name of the Target column.
• mining_imp_val (tuple) – Contains the mining_attributes,mining_strategy, mining_impute_value
• categoric_values (tuple) – Contains Categorical attribute names and its values

Returns:

Nyoka’s Segmentation object

Return type:

segmentation

def get_output(model, target_name)

It returns the output element of the model.

Parameters:

• model – A Scikit-learn model instance.
• target_name (String) – Name of the Target column.

Returns:

Nyoka’s Output object

Return type:

Output

def get_output_for_clustering(values)

Generates output for clustering models

Parameters:model – An instance of Scikit-learn model. Returns:output_fields – Returns Nyoka’s Output object Return type:List

def get_reg_mining_models(model, derived_col_names, col_names, target_name, mining_imp_val, categoric_values, model_name)

Creates xml elements for multi-class linear models

Parameters:

• model – An instance of Scikit-learn model.
• derived_col_names (List) – Contains column names after preprocessing
• col_names (List) – Contains list of feature/column names.
• target_name (String) – Name of the Target column.
• mining_imp_val (tuple) – Contains the mining_attributes,mining_strategy, mining_impute_value
• categoric_values (tuple) – Contains Categorical attribute names and its values
• model_name (string) – Name of the model

Returns:

mining_model – Returns a Nyoka’s MiningModel object

Return type:

List

def get_reg_tab_for_reg_mining_model(model, col_names, index, categorical_values)

Generates Regression Table for multi-class linear models

Parameters:

• model – An instance of Scikit-learn model.
• col_names (List) – Contains list of feature/column names.
• index (int) –
• categoric_values (tuple) – Contains Categorical attribute names and its values

Returns:

Return type:

Returns Nyoka’s RegressionTable object

def get_regrs_models(model, derived_col_names, col_names, target_name, mining_imp_val, categoric_values, model_name)

Generates PMML elements for linear models

Parameters:

• model – A Scikit-learn model instance.
• derived_col_names (List) – Contains column names after preprocessing.
• col_names (List) – Contains list of feature/column names.
• target_name (String) – Name of the Target column.
• mining_imp_val (tuple) – Contains the mining_attributes,mining_strategy, mining_impute_value
• categoric_values (tuple) – Contains Categorical attribute names and its values
• model_name (string) – Name of the model

Returns:

regrs_models – Nyoka’s RegressionModel object

Return type:

List

def get_regrs_tabl(model, feature_names, target_name, categoric_values)

It returns the Regression Table element of the model.

Parameters:

• model – A Scikit-learn model instance.
• derived_col_names (List) – Contains column names after preprocessing.
• target_name (String) – Name of the Target column.
• categoric_values (tuple) – Contains Categorical attribute names and its values

Returns:

merge – Nyoka’s RegressionTable object

Return type:

List

def get_segments(model, derived_col_names, col_names, target_name, mining_imp_val, categoric_values, model_name)

It returns the Segment element of a Segmentation.

Parameters:

• model – A Scikit-learn model instance.
• derived_col_names (List) – Contains column names after preprocessing.
• col_names (List) – Contains list of feature/column names.
• target_name (String) – Name of the Target column.
• mining_imp_val (tuple) – Contains the mining_attributes,mining_strategy, mining_impute_value
• categoric_values (tuple) – Contains Categorical attribute names and its values

Returns:

Nyoka’s Segment object

Return type:

segments

def get_segments_for_gbc(model, derived_col_names, col_names, target_name, mining_imp_val, categoric_values, model_name)

It returns list of Segments element of a Segmentation.

Parameters:

• model – A Scikit-learn model instance.
• derived_col_names (List) – Contains column names after preprocessing.
• col_names (List) – Contains list of feature/column names.
• target_name (String) – Name of the Target column.
• mining_imp_val (tuple) – Contains the mining_attributes,mining_strategy, mining_impute_value
• categoric_values (tuple) – Contains Categorical attribute names and its values

Returns:

segments – Nyoka’s Segment object

Return type:

List

def get_super_cls_names(model_inst)

It returns the set of Super class of the model.

Parameters:model_inst – Instance of the scikit-learn model Returns:parents – Returns all the parent class of the model instance. Return type:Set

def get_supportVectorMachine(model)

Generates PMML elements for support vector machine models

Parameters:model – A Scikit-learn model instance. Returns:support_vector_machines – Nyoka’s SupportVectorMachineModel object Return type:List

def get_supportVectorMachine_models(model, derived_col_names, col_names, target_names, mining_imp_val, categoric_values, model_name)

Generates PMML elements for support vector machine models

Parameters:

• model – An instance of Scikit-learn model.
• derived_col_names (List) – Contains column names after preprocessing.
• col_names (List) – Contains list of feature/column names.
• target_names (String) – Name of the Target column.
• mining_imp_val (tuple) – Contains the mining_attributes,mining_strategy, mining_impute_value
• categoric_values (tuple) – Contains Categorical attribute names and its values
• model_name (string) – Name of the model

Returns:

supportVector_models – Returns Nyoka’s SupportVectorMachineModel object

Return type:

List

def get_targets(model, target_name)

It returns the Target element of the model.

Parameters:

• model – A Scikit-learn model instance.
• target_name (String) – Name of the Target column.

Returns:

Returns Nyoka’s Target object

Return type:

targets

def get_threshold()

It returns the Threshold value for Naive Bayes models.

Returns: Return type:Returns the Threshold value

def get_training_instances(model, derived_col_names, target_name)

It returns the Training Instance element.

Parameters:

• model – An instance of Scikit-learn model.
• derived_col_names (List) – Contains column names after preprocessing
• target_name (String) – Name of the Target column.

Returns:

Returns Nyoka’s TrainingInstances object

Return type:

TrainingInstances

def get_tree_models(model, derived_col_names, col_names, target_name, mining_imp_val, categoric_values, model_name)

Generates PMML elements for tree models

Parameters:

• model – A Scikit-learn model instance.
• derived_col_names – Contains column names after preprocessing.
• col_names (List) – Contains list of feature/column names.
• target_name (String) – Name of the Target column.
• mining_imp_val (tuple) – Contains the mining_attributes,mining_strategy, mining_impute_value
• categoric_values (tuple) – Contains Categorical attribute names and its values
• model_name (string) – Name of the model

Returns:

tree_models – Nyoka’s TreeModel object

Return type:

List

def get_vectorDictionary(model, derived_col_names, categoric_values)

It return the Vector Dictionary element.

Parameters:

• model – A Scikit-learn model instance.
• derived_col_names (List) – Contains column names after preprocessing.
• categoric_values (tuple) – Contains Categorical attribute names and its values

Returns:

Nyoka’s VectorDictionary object

Return type:

VectorDictionary

def has_target(model)

Checks whether a given model has target or not

Parameters:model – Scikit-learn’s model object Returns: Return type:Boolean value

def skl_to_pmml(pipeline, col_names, target_name='target', pmml_f_name='from_sklearn.pmml', model_name=None, description=None)

Exports scikit-learn pipeline object into pmml

Parameters:

• pipeline – Contains an instance of Pipeline with preprocessing and final estimator
• col_names (List) – Contains list of feature/column names.
• target_name (String) – Name of the target column. (Default=’target’)
• pmml_f_name (String) – Name of the pmml file. (Default=’from_sklearn.pmml’)
• model_name (string (optional)) – Name of the model
• description (string (optional)) – Description of the model

Returns:

Return type:

Generates a PMML object and exports it to pmml_f_name