Class: Rumale::Pipeline::Pipeline

Inherits:
Object
  • Object
show all
Includes:
Base::BaseEstimator
Defined in:
lib/rumale/pipeline/pipeline.rb

Overview

Pipeline is a class that implements the function to perform the transformers and estimators sequencially.

Examples:

rbf = Rumale::KernelApproximation::RBF.new(gamma: 1.0, n_coponents: 128, random_seed: 1)
svc = Rumale::LinearModel::SVC.new(reg_param: 1.0, fit_bias: true, max_iter: 5000, random_seed: 1)
pipeline = Rumale::Pipeline::Pipeline.new(steps: { trs: rbf, est: svc })
pipeline.fit(training_samples, traininig_labels)
results = pipeline.predict(testing_samples)

Instance Attribute Summary collapse

Attributes included from Base::BaseEstimator

#params

Instance Method Summary collapse

Constructor Details

#initialize(steps:) ⇒ Pipeline

Create a new pipeline.

Parameters:

  • steps (Hash) —

    List of transformers and estimators. The order of transforms follows the insertion order of hash keys. The last entry is considered an estimator.



30
31
32
33
34
35
 
# File 'lib/rumale/pipeline/pipeline.rb', line 30

def initialize(steps:)
  check_params_type(Hash, steps: steps)
  validate_steps(steps)
  @params = {}
  @steps = steps
end

Instance Attribute Details

#steps ⇒ Hash (readonly)

Return the steps.

Returns:

  • (Hash) 


24
25
26
 
# File 'lib/rumale/pipeline/pipeline.rb', line 24

def steps
  @steps
end

Instance Method Details

#decision_function(x) ⇒ Numo::DFloat

Call the decision_function method of last estimator after applying all transforms.

Parameters:

  • x (Numo::DFloat) —

    (shape: [n_samples, n_features]) The samples to compute the scores.

Returns:

  • (Numo::DFloat) —

    (shape: [n_samples]) Confidence score per sample.



75
76
77
78
79
 
# File 'lib/rumale/pipeline/pipeline.rb', line 75

def decision_function(x)
  check_sample_array(x)
  trans_x = apply_transforms(x)
  last_estimator.decision_function(trans_x)
end

#fit(x, y) ⇒ Pipeline

Fit the model with given training data.

Parameters:

  • x (Numo::DFloat) —

    (shape: [n_samples, n_features]) The training data to be transformed and used for fitting the model.

  • y (Numo::NArray) —

    (shape: [n_samples, n_outputs]) The target values or labels to be used for fitting the model.

Returns:

  • (Pipeline) —

    The learned pipeline itself.



42
43
44
45
46
47
 
# File 'lib/rumale/pipeline/pipeline.rb', line 42

def fit(x, y)
  check_sample_array(x)
  trans_x = apply_transforms(x, y, fit: true)
  last_estimator&.fit(trans_x, y)
  self
end

#fit_predict(x, y = nil) ⇒ Numo::NArray

Call the fit_predict method of last estimator after applying all transforms.

Parameters:

  • x (Numo::DFloat) —

    (shape: [n_samples, n_features]) The training data to be transformed and used for fitting the model.

  • y (Numo::NArray) (defaults to: nil) —

    (shape: [n_samples, n_outputs], default: nil) The target values or labels to be used for fitting the model.

Returns:

  • (Numo::NArray) —

    The predicted results by last estimator.



54
55
56
57
58
 
# File 'lib/rumale/pipeline/pipeline.rb', line 54

def fit_predict(x, y = nil)
  check_sample_array(x)
  trans_x = apply_transforms(x, y, fit: true)
  last_estimator.fit_predict(trans_x)
end

#fit_transform(x, y = nil) ⇒ Numo::NArray

Call the fit_transform method of last estimator after applying all transforms.

Parameters:

  • x (Numo::DFloat) —

    (shape: [n_samples, n_features]) The training data to be transformed and used for fitting the model.

  • y (Numo::NArray) (defaults to: nil) —

    (shape: [n_samples, n_outputs], default: nil) The target values or labels to be used for fitting the model.

Returns:

  • (Numo::NArray) —

    The predicted results by last estimator.



65
66
67
68
69
 
# File 'lib/rumale/pipeline/pipeline.rb', line 65

def fit_transform(x, y = nil)
  check_sample_array(x)
  trans_x = apply_transforms(x, y, fit: true)
  last_estimator.fit_transform(trans_x, y)
end

#inverse_transform(z) ⇒ Numo::DFloat

Call the inverse_transform method in reverse order.

Parameters:

  • z (Numo::DFloat) —

    (shape: [n_samples, n_components]) The transformed samples to be restored into original space.

Returns:

  • (Numo::DFloat) —

    (shape: [n_samples, n_featuress]) The restored samples.



125
126
127
128
129
130
131
132
133
134
 
# File 'lib/rumale/pipeline/pipeline.rb', line 125

def inverse_transform(z)
  check_sample_array(z)
  itrans_z = z
  @steps.keys.reverse_each do |name|
    transformer = @steps[name]
    next if transformer.nil?
    itrans_z = transformer.inverse_transform(itrans_z)
  end
  itrans_z
end

#marshal_dump ⇒ Hash

Dump marshal data.

Returns:

  • (Hash) —

    The marshal data about Pipeline.



149
150
151
152
 
# File 'lib/rumale/pipeline/pipeline.rb', line 149

def marshal_dump
  { params: @params,
    steps: @steps }
end

#marshal_load(obj) ⇒ nil

Load marshal data.

Returns:

  • (nil) 


156
157
158
159
160
 
# File 'lib/rumale/pipeline/pipeline.rb', line 156

def marshal_load(obj)
  @params = obj[:params]
  @steps = obj[:steps]
  nil
end

#predict(x) ⇒ Numo::NArray

Call the predict method of last estimator after applying all transforms.

Parameters:

  • x (Numo::DFloat) —

    (shape: [n_samples, n_features]) The samples to obtain prediction result.

Returns:

  • (Numo::NArray) —

    The predicted results by last estimator.



85
86
87
88
89
 
# File 'lib/rumale/pipeline/pipeline.rb', line 85

def predict(x)
  check_sample_array(x)
  trans_x = apply_transforms(x)
  last_estimator.predict(trans_x)
end

#predict_log_proba(x) ⇒ Numo::DFloat

Call the predict_log_proba method of last estimator after applying all transforms.

Parameters:

  • x (Numo::DFloat) —

    (shape: [n_samples, n_features]) The samples to predict the log-probailities.

Returns:

  • (Numo::DFloat) —

    (shape: [n_samples, n_classes]) Predicted log-probability of each class per sample.



95
96
97
98
99
 
# File 'lib/rumale/pipeline/pipeline.rb', line 95

def predict_log_proba(x)
  check_sample_array(x)
  trans_x = apply_transforms(x)
  last_estimator.predict_log_proba(trans_x)
end

#predict_proba(x) ⇒ Numo::DFloat

Call the predict_proba method of last estimator after applying all transforms.

Parameters:

  • x (Numo::DFloat) —

    (shape: [n_samples, n_features]) The samples to predict the probailities.

Returns:

  • (Numo::DFloat) —

    (shape: [n_samples, n_classes]) Predicted probability of each class per sample.



105
106
107
108
109
 
# File 'lib/rumale/pipeline/pipeline.rb', line 105

def predict_proba(x)
  check_sample_array(x)
  trans_x = apply_transforms(x)
  last_estimator.predict_proba(trans_x)
end

#score(x, y) ⇒ Float

Call the score method of last estimator after applying all transforms.

Parameters:

  • x (Numo::DFloat) —

    (shape: [n_samples, n_features]) Testing data.

  • y (Numo::NArray) —

    (shape: [n_samples, n_outputs]) True target values or labels for testing data.

Returns:

  • (Float) —

    The score of last estimator



141
142
143
144
145
 
# File 'lib/rumale/pipeline/pipeline.rb', line 141

def score(x, y)
  check_sample_array(x)
  trans_x = apply_transforms(x)
  last_estimator.score(trans_x, y)
end

#transform(x) ⇒ Numo::DFloat

Call the transform method of last estimator after applying all transforms.

Parameters:

  • x (Numo::DFloat) —

    (shape: [n_samples, n_features]) The samples to be transformed.

Returns:

  • (Numo::DFloat) —

    (shape: [n_samples, n_components]) The transformed samples.



115
116
117
118
119
 
# File 'lib/rumale/pipeline/pipeline.rb', line 115

def transform(x)
  check_sample_array(x)
  trans_x = apply_transforms(x)
  last_estimator.nil? ? trans_x : last_estimator.transform(trans_x)
end