kernel_pca.rb in rumale-0.18.4

- old
+ new

@@ -9,11 +9,11 @@
     #
     # @example
     #   require 'numo/linalg/autoloader'
     #
     #   kernel_mat_train = Rumale::PairwiseMetric::rbf_kernel(training_samples)
-    #   kpca = Rumale::KernelMachine::KernelPCA(n_components: 2)
+    #   kpca = Rumale::KernelMachine::KernelPCA.new(n_components: 2)
     #   mapped_traininig_samples = kpca.fit_transform(kernel_mat_train)
     #
     #   kernel_mat_test = Rumale::PairwiseMetric::rbf_kernel(test_samples, training_samples)
     #   mapped_test_samples = kpca.transform(kernel_mat_test)
     #
@@ -25,11 +25,11 @@
 
       # Returns the eigenvalues of the centered kernel matrix.
       # @return [Numo::DFloat] (shape: [n_components])
       attr_reader :lambdas
 
-      # Returns the eigenvectros of the centered kernel matrix.
+      # Returns the eigenvectors of the centered kernel matrix.
       # @return [Numo::DFloat] (shape: [n_training_sampes, n_components])
       attr_reader :alphas
 
       # Create a new transformer with Kernel PCA.
       #
@@ -38,10 +38,11 @@
         check_params_numeric(n_components: n_components)
         @params = {}
         @params[:n_components] = n_components
         @alphas = nil
         @lambdas = nil
+        @transform_mat = nil
         @row_mean = nil
         @all_mean = nil
       end
 
       # Fit the model with given training data.
@@ -61,10 +62,11 @@
         @all_mean = @row_mean.sum.fdiv(n_samples)
         centered_kernel_mat = x - x.mean(1).expand_dims(1) - @row_mean + @all_mean
         eig_vals, eig_vecs = Numo::Linalg.eigh(centered_kernel_mat, vals_range: (n_samples - @params[:n_components])...n_samples)
         @alphas = eig_vecs.reverse(1).dup
         @lambdas = eig_vals.reverse.dup
+        @transform_mat = @alphas.dot((1.0 / Numo::NMath.sqrt(@lambdas)).diag)
         self
       end
 
       # Fit the model with training data, and then transform them with the learned model.
       # To execute this method, Numo::Linalg must be loaded.
@@ -85,11 +87,10 @@
       # @return [Numo::DFloat] (shape: [n_testing_samples, n_components]) The transformed data.
       def transform(x)
         x = check_convert_sample_array(x)
         col_mean = x.sum(1) / @row_mean.shape[0]
         centered_kernel_mat = x - col_mean.expand_dims(1) - @row_mean + @all_mean
-        transform_mat = @alphas.dot((1.0 / Numo::NMath.sqrt(@lambdas)).diag)
-        transformed = centered_kernel_mat.dot(transform_mat)
+        transformed = centered_kernel_mat.dot(@transform_mat)
         @params[:n_components] == 1 ? transformed[true, 0].dup : transformed
       end
     end
   end
 end