rumale-svm 0.1.0 → 0.2.0

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (14) hide show
  1. checksums.yaml +4 -4
  2. data/.github/workflows/build.yml +24 -0
  3. data/CHANGELOG.md +7 -0
  4. data/lib/rumale/svm/linear_svc.rb +12 -13
  5. data/lib/rumale/svm/linear_svr.rb +10 -10
  6. data/lib/rumale/svm/logistic_regression.rb +12 -13
  7. data/lib/rumale/svm/nu_svc.rb +15 -15
  8. data/lib/rumale/svm/nu_svr.rb +13 -13
  9. data/lib/rumale/svm/one_class_svm.rb +13 -13
  10. data/lib/rumale/svm/svc.rb +15 -15
  11. data/lib/rumale/svm/svr.rb +15 -14
  12. data/lib/rumale/svm/version.rb +1 -1
  13. data/rumale-svm.gemspec +2 -1
  14. metadata +10 -7
checksums.yaml CHANGED
@@ -1,7 +1,7 @@
1
1
  ---
2
2
  SHA1:
3
- metadata.gz: 40bfeaba6ea5d4b31d75b3c7c7573b121aa2390a
4
- data.tar.gz: a59abc913381101d82c488e145c9e375d2e55f41
3
+ metadata.gz: 94a681354e720d47b0efe86bb2833b4ecdb1c3ff
4
+ data.tar.gz: 050ae3bb322c3a89ebbceb1b5e114988738a9404
5
5
  SHA512:
6
- metadata.gz: 333cce8eb4f15d6a23e7cc31b2cd5d122e3f59ce61971e0fa9d8c01144a4e67b44139a367132752d08c269b7564330503cb13558f93accc37e520f1542688c06
7
- data.tar.gz: c5fc3f8c8287603369c1b801c6615644a13b75fdd75fb21f3470aa2f067d8d0800cac59b9b7fa9984405cf55a8edcf9e8e4ed46f2a65e0bf8e9791372fd5992c
6
+ metadata.gz: 3fe3c396ce596b589b508ecbce62258f14ef3733eb26d9d5829bad81c4bba2380fbf143178e1daabf0d9f523118099a0d56ed285bc5407da1c6af10e56283e23
7
+ data.tar.gz: fa78aa2cb873c8246b9ddc3c75063ae00ba9acf8c644d87b33900e16a813b870f2cffde7159625e695a1c0a3874d48a9302dca93e3670e25d05ba0caa180656c
data/.github/workflows/build.yml ADDED
@@ -0,0 +1,24 @@
1
+ name: build
2
+
3
+ on: [push]
4
+
5
+ jobs:
6
+ build:
7
+
8
+ runs-on: ubuntu-latest
9
+
10
+ strategy:
11
+ matrix:
12
+ ruby: ['2.4.x', '2.5.x', '2.6.x']
13
+
14
+ steps:
15
+ - uses: actions/checkout@master
16
+ - name: Set up Ruby
17
+ uses: actions/setup-ruby@master
18
+ with:
19
+ ruby-version: ${{ matrix.ruby }}
20
+ - name: Build and test with Rake
21
+ run: |
22
+ gem install bundler
23
+ bundle install --jobs 4 --retry 3
24
+ bundle exec rake
data/CHANGELOG.md ADDED
@@ -0,0 +1,7 @@
1
+ # 0.2.0
2
+ - Supported the new Rumale's validation:
3
+ - Fix to use new numeric validation for hyperparameter values.
4
+ - Fix to use new array validation that accepts Ruby Array for samples, labels, and target values.
5
+
6
+ # 0.1.0
7
+ - First release.
data/lib/rumale/svm/linear_svc.rb CHANGED
@@ -42,23 +42,22 @@ module Rumale
42
42
  # @param verbose [Boolean] The flag indicating whether to output learning process message
43
43
  # @param random_seed [Integer/Nil] The seed value using to initialize the random generator.
44
44
  def initialize(penalty: 'l2', loss: 'squared_hinge', dual: true, reg_param: 1.0,
45
- fit_bias: true, bias_scale: 1.0, probability: false,
46
- tol: 1e-3, verbose: false, random_seed: nil)
45
+ fit_bias: true, bias_scale: 1.0, probability: false, tol: 1e-3, verbose: false, random_seed: nil)
47
46
  check_params_string(penalty: penalty, loss: loss)
48
- check_params_float(reg_param: reg_param, bias_scale: bias_scale, tol: tol)
47
+ check_params_numeric(reg_param: reg_param, bias_scale: bias_scale, tol: tol)
49
48
  check_params_boolean(dual: dual, fit_bias: fit_bias, probability: probability, verbose: verbose)
50
- check_params_type_or_nil(Integer, random_seed: random_seed)
49
+ check_params_numeric_or_nil(random_seed: random_seed)
51
50
  @params = {}
52
51
  @params[:penalty] = penalty == 'l1' ? 'l1' : 'l2'
53
52
  @params[:loss] = loss == 'hinge' ? 'hinge' : 'squared_hinge'
54
53
  @params[:dual] = dual
55
- @params[:reg_param] = reg_param
54
+ @params[:reg_param] = reg_param.to_f
56
55
  @params[:fit_bias] = fit_bias
57
- @params[:bias_scale] = bias_scale
56
+ @params[:bias_scale] = bias_scale.to_f
58
57
  @params[:probability] = probability
59
- @params[:tol] = tol
58
+ @params[:tol] = tol.to_f
60
59
  @params[:verbose] = verbose
61
- @params[:random_seed] = random_seed
60
+ @params[:random_seed] = random_seed.nil? ? nil : random_seed.to_i
62
61
  @model = nil
63
62
  @weight_vec = nil
64
63
  @bias_term = nil
@@ -71,8 +70,8 @@ module Rumale
71
70
  # @param y [Numo::Int32] (shape: [n_samples]) The labels to be used for fitting the model.
72
71
  # @return [LinearSVC] The learned classifier itself.
73
72
  def fit(x, y)
74
- check_sample_array(x)
75
- check_label_array(y)
73
+ x = check_convert_sample_array(x)
74
+ y = check_convert_label_array(y)
76
75
  check_sample_label_size(x, y)
77
76
  xx = fit_bias? ? expand_feature(x) : x
78
77
  @model = Numo::Liblinear.train(xx, y, liblinear_params)
@@ -86,7 +85,7 @@ module Rumale
86
85
  # @param x [Numo::DFloat] (shape: [n_samples, n_features]) The samples to compute the scores.
87
86
  # @return [Numo::DFloat] (shape: [n_samples, n_classes]) Confidence score per sample.
88
87
  def decision_function(x)
89
- check_sample_array(x)
88
+ x = check_convert_sample_array(x)
90
89
  xx = fit_bias? ? expand_feature(x) : x
91
90
  Numo::Liblinear.decision_function(xx, liblinear_params, @model)
92
91
  end
@@ -96,7 +95,7 @@ module Rumale
96
95
  # @param x [Numo::DFloat] (shape: [n_samples, n_features]) The samples to predict the labels.
97
96
  # @return [Numo::Int32] (shape: [n_samples]) Predicted class label per sample.
98
97
  def predict(x)
99
- check_sample_array(x)
98
+ x = check_convert_sample_array(x)
100
99
  xx = fit_bias? ? expand_feature(x) : x
101
100
  Numo::Int32.cast(Numo::Liblinear.predict(xx, liblinear_params, @model))
102
101
  end
@@ -107,7 +106,7 @@ module Rumale
107
106
  # @param x [Numo::DFloat] (shape: [n_samples, n_features]) The samples to predict the probailities.
108
107
  # @return [Numo::DFloat] (shape: [n_samples, n_classes]) Predicted probability of each class per sample.
109
108
  def predict_proba(x)
110
- check_sample_array(x)
109
+ x = check_convert_sample_array(x)
111
110
  if binary_class?
112
111
  probs = Numo::DFloat.zeros(x.shape[0], 2)
113
112
  probs[true, 0] = 1.0 / (Numo::NMath.exp(@prob_param[0] * decision_function(x) + @prob_param[1]) + 1.0)
data/lib/rumale/svm/linear_svr.rb CHANGED
@@ -41,19 +41,19 @@ module Rumale
41
41
  def initialize(loss: 'squared_epsilon_insensitive', dual: true, reg_param: 1.0, epsilon: 0.1,
42
42
  fit_bias: true, bias_scale: 1.0, tol: 1e-3, verbose: false, random_seed: nil)
43
43
  check_params_string(loss: loss)
44
- check_params_float(reg_param: reg_param, epsilon: epsilon, bias_scale: bias_scale, tol: tol)
44
+ check_params_numeric(reg_param: reg_param, epsilon: epsilon, bias_scale: bias_scale, tol: tol)
45
45
  check_params_boolean(dual: dual, fit_bias: fit_bias, verbose: verbose)
46
- check_params_type_or_nil(Integer, random_seed: random_seed)
46
+ check_params_numeric_or_nil(random_seed: random_seed)
47
47
  @params = {}
48
48
  @params[:loss] = loss == 'epsilon_insensitive' ? 'epsilon_insensitive' : 'squared_epsilon_insensitive'
49
49
  @params[:dual] = dual
50
- @params[:reg_param] = reg_param
51
- @params[:epsilon] = epsilon
50
+ @params[:reg_param] = reg_param.to_f
51
+ @params[:epsilon] = epsilon.to_f
52
52
  @params[:fit_bias] = fit_bias
53
- @params[:bias_scale] = bias_scale
54
- @params[:tol] = tol
53
+ @params[:bias_scale] = bias_scale.to_f
54
+ @params[:tol] = tol.to_f
55
55
  @params[:verbose] = verbose
56
- @params[:random_seed] = random_seed
56
+ @params[:random_seed] = random_seed.nil? ? nil : random_seed.to_i
57
57
  @model = nil
58
58
  @weight_vec = nil
59
59
  @bias_term = nil
@@ -65,8 +65,8 @@ module Rumale
65
65
  # @param y [Numo::DFloat] (shape: [n_samples]) The target values to be used for fitting the model.
66
66
  # @return [LinearSVR] The learned regressor itself.
67
67
  def fit(x, y)
68
- check_sample_array(x)
69
- check_tvalue_array(y)
68
+ x = check_convert_sample_array(x)
69
+ y = check_convert_tvalue_array(y)
70
70
  check_sample_tvalue_size(x, y)
71
71
  xx = fit_bias? ? expand_feature(x) : x
72
72
  @model = Numo::Liblinear.train(xx, y, liblinear_params)
@@ -79,7 +79,7 @@ module Rumale
79
79
  # @param x [Numo::DFloat] (shape: [n_samples, n_features]) The samples to predict the labels.
80
80
  # @return [Numo::DFloat] (shape: [n_samples]) Predicted value per sample.
81
81
  def predict(x)
82
- check_sample_array(x)
82
+ x = check_convert_sample_array(x)
83
83
  xx = fit_bias? ? expand_feature(x) : x
84
84
  Numo::Liblinear.predict(xx, liblinear_params, @model)
85
85
  end
data/lib/rumale/svm/logistic_regression.rb CHANGED
@@ -38,21 +38,20 @@ module Rumale
38
38
  # @param verbose [Boolean] The flag indicating whether to output learning process message
39
39
  # @param random_seed [Integer/Nil] The seed value using to initialize the random generator.
40
40
  def initialize(penalty: 'l2', dual: true, reg_param: 1.0,
41
- fit_bias: true, bias_scale: 1.0,
42
- tol: 1e-3, verbose: false, random_seed: nil)
41
+ fit_bias: true, bias_scale: 1.0, tol: 1e-3, verbose: false, random_seed: nil)
43
42
  check_params_string(penalty: penalty)
44
- check_params_float(reg_param: reg_param, bias_scale: bias_scale, tol: tol)
43
+ check_params_numeric(reg_param: reg_param, bias_scale: bias_scale, tol: tol)
45
44
  check_params_boolean(dual: dual, fit_bias: fit_bias, verbose: verbose)
46
- check_params_type_or_nil(Integer, random_seed: random_seed)
45
+ check_params_numeric_or_nil(random_seed: random_seed)
47
46
  @params = {}
48
47
  @params[:penalty] = penalty == 'l1' ? 'l1' : 'l2'
49
48
  @params[:dual] = dual
50
- @params[:reg_param] = reg_param
49
+ @params[:reg_param] = reg_param.to_f
51
50
  @params[:fit_bias] = fit_bias
52
- @params[:bias_scale] = bias_scale
53
- @params[:tol] = tol
51
+ @params[:bias_scale] = bias_scale.to_f
52
+ @params[:tol] = tol.to_f
54
53
  @params[:verbose] = verbose
55
- @params[:random_seed] = random_seed
54
+ @params[:random_seed] = random_seed.nil? ? nil : random_seed.to_i
56
55
  @model = nil
57
56
  @weight_vec = nil
58
57
  @bias_term = nil
@@ -64,8 +63,8 @@ module Rumale
64
63
  # @param y [Numo::Int32] (shape: [n_samples]) The labels to be used for fitting the model.
65
64
  # @return [LogisticRegression] The learned classifier itself.
66
65
  def fit(x, y)
67
- check_sample_array(x)
68
- check_label_array(y)
66
+ x = check_convert_sample_array(x)
67
+ y = check_convert_label_array(y)
69
68
  check_sample_label_size(x, y)
70
69
  xx = fit_bias? ? expand_feature(x) : x
71
70
  @model = Numo::Liblinear.train(xx, y, liblinear_params)
@@ -78,7 +77,7 @@ module Rumale
78
77
  # @param x [Numo::DFloat] (shape: [n_samples, n_features]) The samples to compute the scores.
79
78
  # @return [Numo::DFloat] (shape: [n_samples, n_classes]) Confidence score per sample.
80
79
  def decision_function(x)
81
- check_sample_array(x)
80
+ x = check_convert_sample_array(x)
82
81
  xx = fit_bias? ? expand_feature(x) : x
83
82
  Numo::Liblinear.decision_function(xx, liblinear_params, @model)
84
83
  end
@@ -88,7 +87,7 @@ module Rumale
88
87
  # @param x [Numo::DFloat] (shape: [n_samples, n_features]) The samples to predict the labels.
89
88
  # @return [Numo::Int32] (shape: [n_samples]) Predicted class label per sample.
90
89
  def predict(x)
91
- check_sample_array(x)
90
+ x = check_convert_sample_array(x)
92
91
  xx = fit_bias? ? expand_feature(x) : x
93
92
  Numo::Int32.cast(Numo::Liblinear.predict(xx, liblinear_params, @model))
94
93
  end
@@ -99,7 +98,7 @@ module Rumale
99
98
  # @param x [Numo::DFloat] (shape: [n_samples, n_features]) The samples to predict the probailities.
100
99
  # @return [Numo::DFloat] (shape: [n_samples, n_classes]) Predicted probability of each class per sample.
101
100
  def predict_proba(x)
102
- check_sample_array(x)
101
+ x = check_convert_sample_array(x)
103
102
  xx = fit_bias? ? expand_feature(x) : x
104
103
  Numo::Liblinear.predict_proba(xx, liblinear_params, @model)
105
104
  end
data/lib/rumale/svm/nu_svc.rb CHANGED
@@ -31,22 +31,22 @@ module Rumale
31
31
  # @param random_seed [Integer/Nil] The seed value using to initialize the random generator.
32
32
  def initialize(nu: 0.5, kernel: 'rbf', degree: 3, gamma: 1.0, coef0: 0.0,
33
33
  shrinking: true, probability: true, cache_size: 200.0, tol: 1e-3, verbose: false, random_seed: nil)
34
- check_params_float(nu: nu, gamma: gamma, coef0: coef0, cache_size: cache_size, tol: tol)
35
- check_params_integer(degree: degree)
34
+ check_params_numeric(nu: nu, degree: degree, gamma: gamma, coef0: coef0, cache_size: cache_size, tol: tol)
35
+ check_params_string(kernel: kernel)
36
36
  check_params_boolean(shrinking: shrinking, probability: probability, verbose: verbose)
37
- check_params_type_or_nil(Integer, random_seed: random_seed)
37
+ check_params_numeric_or_nil(random_seed: random_seed)
38
38
  @params = {}
39
- @params[:nu] = nu
39
+ @params[:nu] = nu.to_f
40
40
  @params[:kernel] = kernel
41
- @params[:degree] = degree
42
- @params[:gamma] = gamma
43
- @params[:coef0] = coef0
41
+ @params[:degree] = degree.to_i
42
+ @params[:gamma] = gamma.to_f
43
+ @params[:coef0] = coef0.to_f
44
44
  @params[:shrinking] = shrinking
45
45
  @params[:probability] = probability
46
- @params[:cache_size] = cache_size
47
- @params[:tol] = tol
46
+ @params[:cache_size] = cache_size.to_f
47
+ @params[:tol] = tol.to_f
48
48
  @params[:verbose] = verbose
49
- @params[:random_seed] = random_seed
49
+ @params[:random_seed] = random_seed.nil? ? nil : random_seed.to_i
50
50
  @model = nil
51
51
  end
52
52
 
@@ -57,8 +57,8 @@ module Rumale
57
57
  # @param y [Numo::Int32] (shape: [n_samples]) The labels to be used for fitting the model.
58
58
  # @return [NuSVC] The learned classifier itself.
59
59
  def fit(x, y)
60
- check_sample_array(x)
61
- check_label_array(y)
60
+ x = check_convert_sample_array(x)
61
+ y = check_convert_label_array(y)
62
62
  check_sample_label_size(x, y)
63
63
  xx = precomputed_kernel? ? add_index_col(x) : x
64
64
  @model = Numo::Libsvm.train(xx, y, libsvm_params)
@@ -71,7 +71,7 @@ module Rumale
71
71
  # If the kernel is 'precomputed', the shape of x must be [n_samples, n_training_samples].
72
72
  # @return [Numo::DFloat] (shape: [n_samples, n_classes]) Confidence score per sample.
73
73
  def decision_function(x)
74
- check_sample_array(x)
74
+ x = check_convert_sample_array(x)
75
75
  xx = precomputed_kernel? ? add_index_col(x) : x
76
76
  Numo::Libsvm.decision_function(xx, libsvm_params, @model)
77
77
  end
@@ -82,7 +82,7 @@ module Rumale
82
82
  # If the kernel is 'precomputed', the shape of x must be [n_samples, n_training_samples].
83
83
  # @return [Numo::Int32] (shape: [n_samples]) Predicted class label per sample.
84
84
  def predict(x)
85
- check_sample_array(x)
85
+ x = check_convert_sample_array(x)
86
86
  xx = precomputed_kernel? ? add_index_col(x) : x
87
87
  Numo::Int32.cast(Numo::Libsvm.predict(xx, libsvm_params, @model))
88
88
  end
@@ -94,7 +94,7 @@ module Rumale
94
94
  # If the kernel is 'precomputed', the shape of x must be [n_samples, n_training_samples].
95
95
  # @return [Numo::DFloat] (shape: [n_samples, n_classes]) Predicted probability of each class per sample.
96
96
  def predict_proba(x)
97
- check_sample_array(x)
97
+ x = check_convert_sample_array(x)
98
98
  xx = precomputed_kernel? ? add_index_col(x) : x
99
99
  Numo::Libsvm.predict_proba(xx, libsvm_params, @model)
100
100
  end
data/lib/rumale/svm/nu_svr.rb CHANGED
@@ -30,21 +30,21 @@ module Rumale
30
30
  # @param random_seed [Integer/Nil] The seed value using to initialize the random generator.
31
31
  def initialize(nu: 0.5, kernel: 'rbf', degree: 3, gamma: 1.0, coef0: 0.0,
32
32
  shrinking: true, cache_size: 200.0, tol: 1e-3, verbose: false, random_seed: nil)
33
- check_params_float(nu: nu, gamma: gamma, coef0: coef0, cache_size: cache_size, tol: tol)
34
- check_params_integer(degree: degree)
33
+ check_params_numeric(nu: nu, degree: degree, gamma: gamma, coef0: coef0, cache_size: cache_size, tol: tol)
34
+ check_params_string(kernel: kernel)
35
35
  check_params_boolean(shrinking: shrinking, verbose: verbose)
36
- check_params_type_or_nil(Integer, random_seed: random_seed)
36
+ check_params_numeric_or_nil(random_seed: random_seed)
37
37
  @params = {}
38
- @params[:nu] = nu
38
+ @params[:nu] = nu.to_f
39
39
  @params[:kernel] = kernel
40
- @params[:degree] = degree
41
- @params[:gamma] = gamma
42
- @params[:coef0] = coef0
40
+ @params[:degree] = degree.to_i
41
+ @params[:gamma] = gamma.to_f
42
+ @params[:coef0] = coef0.to_f
43
43
  @params[:shrinking] = shrinking
44
- @params[:cache_size] = cache_size
45
- @params[:tol] = tol
44
+ @params[:cache_size] = cache_size.to_f
45
+ @params[:tol] = tol.to_f
46
46
  @params[:verbose] = verbose
47
- @params[:random_seed] = random_seed
47
+ @params[:random_seed] = random_seed.nil? ? nil : random_seed.to_i
48
48
  @model = nil
49
49
  end
50
50
 
@@ -55,8 +55,8 @@ module Rumale
55
55
  # @param y [Numo::DFloat] (shape: [n_samples]) The target values to be used for fitting the model.
56
56
  # @return [NuSVR] The learned regressor itself.
57
57
  def fit(x, y)
58
- check_sample_array(x)
59
- check_tvalue_array(y)
58
+ x = check_convert_sample_array(x)
59
+ y = check_convert_tvalue_array(y)
60
60
  check_sample_tvalue_size(x, y)
61
61
  xx = precomputed_kernel? ? add_index_col(x) : x
62
62
  @model = Numo::Libsvm.train(xx, y, libsvm_params)
@@ -69,7 +69,7 @@ module Rumale
69
69
  # If the kernel is 'precomputed', the shape of x must be [n_samples, n_training_samples].
70
70
  # @return [Numo::DFloat] (shape: [n_samples]) Predicted value per sample.
71
71
  def predict(x)
72
- check_sample_array(x)
72
+ x = check_convert_sample_array(x)
73
73
  xx = precomputed_kernel? ? add_index_col(x) : x
74
74
  Numo::Libsvm.predict(xx, libsvm_params, @model)
75
75
  end
data/lib/rumale/svm/one_class_svm.rb CHANGED
@@ -30,21 +30,21 @@ module Rumale
30
30
  # @param random_seed [Integer/Nil] The seed value using to initialize the random generator.
31
31
  def initialize(nu: 1.0, kernel: 'rbf', degree: 3, gamma: 1.0, coef0: 0.0,
32
32
  shrinking: true, cache_size: 200.0, tol: 1e-3, verbose: false, random_seed: nil)
33
- check_params_float(nu: nu, gamma: gamma, coef0: coef0, cache_size: cache_size, tol: tol)
34
- check_params_integer(degree: degree)
33
+ check_params_numeric(nu: nu, degree: degree, gamma: gamma, coef0: coef0, cache_size: cache_size, tol: tol)
34
+ check_params_string(kernel: kernel)
35
35
  check_params_boolean(shrinking: shrinking, verbose: verbose)
36
- check_params_type_or_nil(Integer, random_seed: random_seed)
36
+ check_params_numeric_or_nil(random_seed: random_seed)
37
37
  @params = {}
38
- @params[:nu] = nu
38
+ @params[:nu] = nu.to_f
39
39
  @params[:kernel] = kernel
40
- @params[:degree] = degree
41
- @params[:gamma] = gamma
42
- @params[:coef0] = coef0
40
+ @params[:degree] = degree.to_i
41
+ @params[:gamma] = gamma.to_f
42
+ @params[:coef0] = coef0.to_f
43
43
  @params[:shrinking] = shrinking
44
- @params[:cache_size] = cache_size
45
- @params[:tol] = tol
44
+ @params[:cache_size] = cache_size.to_f
45
+ @params[:tol] = tol.to_f
46
46
  @params[:verbose] = verbose
47
- @params[:random_seed] = random_seed
47
+ @params[:random_seed] = random_seed.nil? ? nil : random_seed.to_i
48
48
  @model = nil
49
49
  end
50
50
 
@@ -55,7 +55,7 @@ module Rumale
55
55
  # If the kernel is 'precomputed', x must be a square distance matrix (shape: [n_samples, n_samples]).
56
56
  # @return [OneClassSVM] The learned estimator itself.
57
57
  def fit(x, _y = nil)
58
- check_sample_array(x)
58
+ x = check_convert_sample_array(x)
59
59
  dummy = Numo::DFloat.ones(x.shape[0])
60
60
  @model = Numo::Libsvm.train(x, dummy, libsvm_params)
61
61
  self
@@ -67,7 +67,7 @@ module Rumale
67
67
  # If the kernel is 'precomputed', the shape of x must be [n_samples, n_training_samples].
68
68
  # @return [Numo::DFloat] (shape: [n_samples, n_classes]) Confidence score per sample.
69
69
  def decision_function(x)
70
- check_sample_array(x)
70
+ x = check_convert_sample_array(x)
71
71
  Numo::Libsvm.decision_function(x, libsvm_params, @model)
72
72
  end
73
73
 
@@ -77,7 +77,7 @@ module Rumale
77
77
  # If the kernel is 'precomputed', the shape of x must be [n_samples, n_training_samples].
78
78
  # @return [Numo::Int32] (shape: [n_samples]) Predicted label per sample.
79
79
  def predict(x)
80
- check_sample_array(x)
80
+ x = check_convert_sample_array(x)
81
81
  Numo::Int32.cast(Numo::Libsvm.predict(x, libsvm_params, @model))
82
82
  end
83
83
 
data/lib/rumale/svm/svc.rb CHANGED
@@ -31,22 +31,22 @@ module Rumale
31
31
  # @param random_seed [Integer/Nil] The seed value using to initialize the random generator.
32
32
  def initialize(reg_param: 1.0, kernel: 'rbf', degree: 3, gamma: 1.0, coef0: 0.0,
33
33
  shrinking: true, probability: true, cache_size: 200.0, tol: 1e-3, verbose: false, random_seed: nil)
34
- check_params_float(reg_param: reg_param, gamma: gamma, coef0: coef0, cache_size: cache_size, tol: tol)
35
- check_params_integer(degree: degree)
34
+ check_params_numeric(reg_param: reg_param, degree: degree, gamma: gamma, coef0: coef0, cache_size: cache_size, tol: tol)
35
+ check_params_string(kernel: kernel)
36
36
  check_params_boolean(shrinking: shrinking, probability: probability, verbose: verbose)
37
- check_params_type_or_nil(Integer, random_seed: random_seed)
37
+ check_params_numeric_or_nil(random_seed: random_seed)
38
38
  @params = {}
39
- @params[:reg_param] = reg_param
39
+ @params[:reg_param] = reg_param.to_f
40
40
  @params[:kernel] = kernel
41
- @params[:degree] = degree
42
- @params[:gamma] = gamma
43
- @params[:coef0] = coef0
41
+ @params[:degree] = degree.to_i
42
+ @params[:gamma] = gamma.to_f
43
+ @params[:coef0] = coef0.to_f
44
44
  @params[:shrinking] = shrinking
45
45
  @params[:probability] = probability
46
- @params[:cache_size] = cache_size
47
- @params[:tol] = tol
46
+ @params[:cache_size] = cache_size.to_f
47
+ @params[:tol] = tol.to_f
48
48
  @params[:verbose] = verbose
49
- @params[:random_seed] = random_seed
49
+ @params[:random_seed] = random_seed.nil? ? nil : random_seed.to_i
50
50
  @model = nil
51
51
  end
52
52
 
@@ -57,8 +57,8 @@ module Rumale
57
57
  # @param y [Numo::Int32] (shape: [n_samples]) The labels to be used for fitting the model.
58
58
  # @return [SVC] The learned classifier itself.
59
59
  def fit(x, y)
60
- check_sample_array(x)
61
- check_label_array(y)
60
+ x = check_convert_sample_array(x)
61
+ y = check_convert_label_array(y)
62
62
  check_sample_label_size(x, y)
63
63
  xx = precomputed_kernel? ? add_index_col(x) : x
64
64
  @model = Numo::Libsvm.train(xx, y, libsvm_params)
@@ -71,7 +71,7 @@ module Rumale
71
71
  # If the kernel is 'precomputed', the shape of x must be [n_samples, n_training_samples].
72
72
  # @return [Numo::DFloat] (shape: [n_samples, n_classes]) Confidence score per sample.
73
73
  def decision_function(x)
74
- check_sample_array(x)
74
+ x = check_convert_sample_array(x)
75
75
  xx = precomputed_kernel? ? add_index_col(x) : x
76
76
  Numo::Libsvm.decision_function(xx, libsvm_params, @model)
77
77
  end
@@ -82,7 +82,7 @@ module Rumale
82
82
  # If the kernel is 'precomputed', the shape of x must be [n_samples, n_training_samples].
83
83
  # @return [Numo::Int32] (shape: [n_samples]) Predicted class label per sample.
84
84
  def predict(x)
85
- check_sample_array(x)
85
+ x = check_convert_sample_array(x)
86
86
  xx = precomputed_kernel? ? add_index_col(x) : x
87
87
  Numo::Int32.cast(Numo::Libsvm.predict(xx, libsvm_params, @model))
88
88
  end
@@ -94,7 +94,7 @@ module Rumale
94
94
  # If the kernel is 'precomputed', the shape of x must be [n_samples, n_training_samples].
95
95
  # @return [Numo::DFloat] (shape: [n_samples, n_classes]) Predicted probability of each class per sample.
96
96
  def predict_proba(x)
97
- check_sample_array(x)
97
+ x = check_convert_sample_array(x)
98
98
  xx = precomputed_kernel? ? add_index_col(x) : x
99
99
  Numo::Libsvm.predict_proba(xx, libsvm_params, @model)
100
100
  end
data/lib/rumale/svm/svr.rb CHANGED
@@ -31,22 +31,23 @@ module Rumale
31
31
  # @param random_seed [Integer/Nil] The seed value using to initialize the random generator.
32
32
  def initialize(reg_param: 1.0, epsilon: 0.1, kernel: 'rbf', degree: 3, gamma: 1.0, coef0: 0.0,
33
33
  shrinking: true, cache_size: 200.0, tol: 1e-3, verbose: false, random_seed: nil)
34
- check_params_float(reg_param: reg_param, epsilon: epsilon, gamma: gamma, coef0: coef0, cache_size: cache_size, tol: tol)
35
- check_params_integer(degree: degree)
34
+ check_params_numeric(reg_param: reg_param, degree: degree, epsilon: epsilon, gamma: gamma, coef0: coef0,
35
+ cache_size: cache_size, tol: tol)
36
+ check_params_string(kernel: kernel)
36
37
  check_params_boolean(shrinking: shrinking, verbose: verbose)
37
- check_params_type_or_nil(Integer, random_seed: random_seed)
38
+ check_params_numeric_or_nil(random_seed: random_seed)
38
39
  @params = {}
39
- @params[:reg_param] = reg_param
40
- @params[:epsilon] = epsilon
40
+ @params[:reg_param] = reg_param.to_f
41
+ @params[:epsilon] = epsilon.to_f
41
42
  @params[:kernel] = kernel
42
- @params[:degree] = degree
43
- @params[:gamma] = gamma
44
- @params[:coef0] = coef0
43
+ @params[:degree] = degree.to_i
44
+ @params[:gamma] = gamma.to_f
45
+ @params[:coef0] = coef0.to_f
45
46
  @params[:shrinking] = shrinking
46
- @params[:cache_size] = cache_size
47
- @params[:tol] = tol
47
+ @params[:cache_size] = cache_size.to_f
48
+ @params[:tol] = tol.to_f
48
49
  @params[:verbose] = verbose
49
- @params[:random_seed] = random_seed
50
+ @params[:random_seed] = random_seed.nil? ? nil : random_seed.to_i
50
51
  @model = nil
51
52
  end
52
53
 
@@ -57,8 +58,8 @@ module Rumale
57
58
  # @param y [Numo::DFloat] (shape: [n_samples]) The target values to be used for fitting the model.
58
59
  # @return [SVR] The learned regressor itself.
59
60
  def fit(x, y)
60
- check_sample_array(x)
61
- check_tvalue_array(y)
61
+ x = check_convert_sample_array(x)
62
+ y = check_convert_tvalue_array(y)
62
63
  check_sample_tvalue_size(x, y)
63
64
  xx = precomputed_kernel? ? add_index_col(x) : x
64
65
  @model = Numo::Libsvm.train(xx, y, libsvm_params)
@@ -71,7 +72,7 @@ module Rumale
71
72
  # If the kernel is 'precomputed', the shape of x must be [n_samples, n_training_samples].
72
73
  # @return [Numo::DFloat] (shape: [n_samples]) Predicted value per sample.
73
74
  def predict(x)
74
- check_sample_array(x)
75
+ x = check_convert_sample_array(x)
75
76
  xx = precomputed_kernel? ? add_index_col(x) : x
76
77
  Numo::Libsvm.predict(xx, libsvm_params, @model)
77
78
  end
data/lib/rumale/svm/version.rb CHANGED
@@ -5,6 +5,6 @@ module Rumale
5
5
  # This module consists of Rumale interfaces for suppor vector machine algorithms with LIBSVM and LIBLINEAR.
6
6
  module SVM
7
7
  # The version of Rumale-SVM you are using.
8
- VERSION = '0.1.0'
8
+ VERSION = '0.2.0'
9
9
  end
10
10
  end
data/rumale-svm.gemspec CHANGED
@@ -15,6 +15,7 @@ Gem::Specification.new do |spec|
15
15
  Rumale-SVM provides support vector machine algorithms of LIBSVM and LIBLINEAR with Rumale interface.
16
16
  MSG
17
17
  spec.homepage = 'https://github.com/yoshoku/rumale-svm'
18
+ spec.license = 'BSD-3-Clause'
18
19
 
19
20
  spec.metadata['homepage_uri'] = spec.homepage
20
21
  spec.metadata['source_code_uri'] = 'https://github.com/yoshoku/rumale-svm'
@@ -32,7 +33,7 @@ Gem::Specification.new do |spec|
32
33
 
33
34
  spec.add_runtime_dependency 'numo-liblinear', '~> 1.0'
34
35
  spec.add_runtime_dependency 'numo-libsvm', '~> 1.0'
35
- spec.add_runtime_dependency 'rumale'
36
+ spec.add_runtime_dependency 'rumale', '~> 0.14'
36
37
  spec.add_development_dependency 'bundler', '~> 2.0'
37
38
  spec.add_development_dependency 'coveralls', '~> 0.8'
38
39
  spec.add_development_dependency 'rake', '~> 10.0'
metadata CHANGED
@@ -1,14 +1,14 @@
1
1
  --- !ruby/object:Gem::Specification
2
2
  name: rumale-svm
3
3
  version: !ruby/object:Gem::Version
4
- version: 0.1.0
4
+ version: 0.2.0
5
5
  platform: ruby
6
6
  authors:
7
7
  - yoshoku
8
8
  autorequire:
9
9
  bindir: exe
10
10
  cert_chain: []
11
- date: 2019-10-22 00:00:00.000000000 Z
11
+ date: 2019-11-18 00:00:00.000000000 Z
12
12
  dependencies:
13
13
  - !ruby/object:Gem::Dependency
14
14
  name: numo-liblinear
@@ -42,16 +42,16 @@ dependencies:
42
42
  name: rumale
43
43
  requirement: !ruby/object:Gem::Requirement
44
44
  requirements:
45
- - - ">="
45
+ - - "~>"
46
46
  - !ruby/object:Gem::Version
47
- version: '0'
47
+ version: '0.14'
48
48
  type: :runtime
49
49
  prerelease: false
50
50
  version_requirements: !ruby/object:Gem::Requirement
51
51
  requirements:
52
- - - ">="
52
+ - - "~>"
53
53
  - !ruby/object:Gem::Version
54
- version: '0'
54
+ version: '0.14'
55
55
  - !ruby/object:Gem::Dependency
56
56
  name: bundler
57
57
  requirement: !ruby/object:Gem::Requirement
@@ -119,9 +119,11 @@ extensions: []
119
119
  extra_rdoc_files: []
120
120
  files:
121
121
  - ".coveralls.yml"
122
+ - ".github/workflows/build.yml"
122
123
  - ".gitignore"
123
124
  - ".rspec"
124
125
  - ".travis.yml"
126
+ - CHANGELOG.md
125
127
  - CODE_OF_CONDUCT.md
126
128
  - Gemfile
127
129
  - LICENSE.txt
@@ -141,7 +143,8 @@ files:
141
143
  - lib/rumale/svm/version.rb
142
144
  - rumale-svm.gemspec
143
145
  homepage: https://github.com/yoshoku/rumale-svm
144
- licenses: []
146
+ licenses:
147
+ - BSD-3-Clause
145
148
  metadata:
146
149
  homepage_uri: https://github.com/yoshoku/rumale-svm
147
150
  source_code_uri: https://github.com/yoshoku/rumale-svm