MISC renamed n_iterations to n_iter in all other places.

Author: amueller

benchmarks/bench_plot_svd.py

@@ -12,7 +12,7 @@
 from sklearn.datasets.samples_generator import make_low_rank_matrix
 
 
-def compute_bench(samples_range, features_range, n_iterations=3, rank=50):
+def compute_bench(samples_range, features_range, n_iter=3, rank=50):
 
     it = 0
 
@@ -36,19 +36,19 @@ def compute_bench(samples_range, features_range, n_iterations=3, rank=50):
             results['scipy svd'].append(time() - tstart)
 
             gc.collect()
-            print "benching scikit-learn randomized_svd: n_iterations=0"
+            print "benching scikit-learn randomized_svd: n_iter=0"
             tstart = time()
-            randomized_svd(X, rank, n_iterations=0)
-            results['scikit-learn randomized_svd (n_iterations=0)'].append(
+            randomized_svd(X, rank, n_iter=0)
+            results['scikit-learn randomized_svd (n_iter=0)'].append(
                 time() - tstart)
 
             gc.collect()
-            print ("benching scikit-learn randomized_svd: n_iterations=%d "
-                   % n_iterations)
+            print ("benching scikit-learn randomized_svd: n_iter=%d "
+                   % n_iter)
             tstart = time()
-            randomized_svd(X, rank, n_iterations=n_iterations)
-            results['scikit-learn randomized_svd (n_iterations=%d)'
-                    % n_iterations].append(time() - tstart)
+            randomized_svd(X, rank, n_iter=n_iter)
+            results['scikit-learn randomized_svd (n_iter=%d)'
+                    % n_iter].append(time() - tstart)
 
     return results
 

examples/applications/wikipedia_principal_eigenvector.py

@@ -172,7 +172,7 @@ def get_adjacency_matrix(redirects_filename, page_links_filename, limit=None):
 
 print "Computing the principal singular vectors using randomized_svd"
 t0 = time()
-U, s, V = randomized_svd(X, 5, n_iterations=3)
+U, s, V = randomized_svd(X, 5, n_iter=3)
 print "done in %0.3fs" % (time() - t0)
 
 # print the names of the wikipedia related strongest compenents of the the

examples/svm/plot_svm_scale_c.py

@@ -128,7 +128,7 @@
         # reduce the variance
         grid = GridSearchCV(clf, refit=False, param_grid=param_grid,
                         cv=ShuffleSplit(n=n_samples, train_size=train_size,
-                                        n_iterations=250, random_state=1))
+                                        n_iter=250, random_state=1))
         grid.fit(X, y)
         scores = [x[1] for x in grid.grid_scores_]
 

sklearn/cluster/k_means_.py

@@ -897,7 +897,7 @@ def _mini_batch_step(X, x_squared_norms, centers, counts,
     return inertia, squared_diff
 
 
-def _mini_batch_convergence(model, iteration_idx, n_iterations, tol,
+def _mini_batch_convergence(model, iteration_idx, n_iter, tol,
                             n_samples, centers_squared_diff, batch_inertia,
                             context, verbose=0):
     """Helper function to encapsulte the early stopping logic"""
@@ -926,7 +926,7 @@ def _mini_batch_convergence(model, iteration_idx, n_iterations, tol,
         progress_msg = (
             'Minibatch iteration %d/%d:'
             'mean batch inertia: %f, ewa inertia: %f ' % (
-                iteration_idx + 1, n_iterations, batch_inertia,
+                iteration_idx + 1, n_iter, batch_inertia,
                 ewa_inertia))
         print progress_msg
 
@@ -935,7 +935,7 @@ def _mini_batch_convergence(model, iteration_idx, n_iterations, tol,
     if tol > 0.0 and ewa_diff < tol:
         if verbose:
             print 'Converged (small centers change) at iteration %d/%d' % (
-                iteration_idx + 1, n_iterations)
+                iteration_idx + 1, n_iter)
         return True
 
     # Early stopping heuristic due to lack of improvement on smoothed inertia
@@ -952,7 +952,7 @@ def _mini_batch_convergence(model, iteration_idx, n_iterations, tol,
         if verbose:
             print ('Converged (lack of improvement in inertia)'
                    ' at iteration %d/%d' % (
-                       iteration_idx + 1, n_iterations))
+                       iteration_idx + 1, n_iter))
         return True
 
     # update the convergence context to maintain state across sucessive calls:
@@ -1102,7 +1102,7 @@ def fit(self, X, y=None):
 
         distances = np.zeros(self.batch_size, dtype=np.float64)
         n_batches = int(np.ceil(float(n_samples) / self.batch_size))
-        n_iterations = int(self.max_iter * n_batches)
+        n_iter = int(self.max_iter * n_batches)
 
         init_size = self.init_size
         if init_size is None:
@@ -1158,7 +1158,7 @@ def fit(self, X, y=None):
 
         # Perform the iterative optimization untill the final convergence
         # criterion
-        for iteration_idx in xrange(n_iterations):
+        for iteration_idx in xrange(n_iter):
 
             # Sample the minibatch from the full dataset
             minibatch_indices = self.random_state.random_integers(
@@ -1172,7 +1172,7 @@ def fit(self, X, y=None):
 
             # Monitor the convergence and do early stopping if necessary
             if _mini_batch_convergence(
-                self, iteration_idx, n_iterations, tol, n_samples,
+                self, iteration_idx, n_iter, tol, n_samples,
                 centers_squared_diff, batch_inertia, convergence_context,
                 verbose=self.verbose):
                 break

sklearn/decomposition/pca.py

@@ -485,7 +485,7 @@ def fit(self, X, y=None):
             n_components = self.n_components
 
         U, S, V = randomized_svd(X, n_components,
-                                 n_iterations=self.iterated_power,
+                                 n_iter=self.iterated_power,
                                  random_state=self.random_state)
 
         self.explained_variance_ = exp_var = (S ** 2) / n_samples

sklearn/utils/extmath.py

@@ -4,6 +4,7 @@
 # Authors: G. Varoquaux, A. Gramfort, A. Passos, O. Grisel
 # License: BSD
 
+import warnings
 import numpy as np
 from scipy import linalg
 
@@ -78,7 +79,8 @@ def safe_sparse_dot(a, b, dense_output=False):
         return np.dot(a, b)
 
 
-def randomized_range_finder(A, size, n_iterations, random_state=None):
+def randomized_range_finder(A, size, n_iter, random_state=None,
+        n_iterations=None):
     """Computes an orthonormal matrix whose range approximates the range of A.
 
     Parameters
@@ -87,7 +89,7 @@ def randomized_range_finder(A, size, n_iterations, random_state=None):
         The input data matrix
     size: integer
         Size of the return array
-    n_iterations: integer
+    n_iter: integer
         Number of power iterations used to stabilize the result
     random_state: RandomState or an int seed (0 by default)
         A random number generator instance
@@ -106,6 +108,10 @@ def randomized_range_finder(A, size, n_iterations, random_state=None):
     approximate matrix decompositions
     Halko, et al., 2009 (arXiv:909) http://arxiv.org/pdf/0909.4061
     """
+    if n_iterations is not None:
+        warnings.warn("n_iterations was renamed to n_iter for consistency "
+                "and will be removed in 0.16.", DeprecationWarning)
+        n_iter = n_iterations
     random_state = check_random_state(random_state)
 
     # generating random gaussian vectors r with shape: (A.shape[1], size)
@@ -117,16 +123,16 @@ def randomized_range_finder(A, size, n_iterations, random_state=None):
 
     # perform power iterations with Y to further 'imprint' the top
     # singular vectors of A in Y
-    for i in xrange(n_iterations):
+    for i in xrange(n_iter):
         Y = safe_sparse_dot(A, safe_sparse_dot(A.T, Y))
 
     # extracting an orthonormal basis of the A range samples
     Q, R = qr_economic(Y)
     return Q
 
 
-def randomized_svd(M, n_components, n_oversamples=10, n_iterations=0,
-                   transpose='auto', random_state=0):
+def randomized_svd(M, n_components, n_oversamples=10, n_iter=0,
+                   transpose='auto', random_state=0, n_iterations=None):
     """Computes a truncated randomized SVD
 
     Parameters
@@ -142,7 +148,7 @@ def randomized_svd(M, n_components, n_oversamples=10, n_iterations=0,
         to ensure proper conditioning. The total number of random vectors
         used to find the range of M is n_components + n_oversamples.
 
-    n_iterations: int (default is 0)
+    n_iter: int (default is 0)
         Number of power iterations (can be used to deal with very noisy
         problems).
 
@@ -172,6 +178,11 @@ def randomized_svd(M, n_components, n_oversamples=10, n_iterations=0,
     * A randomized algorithm for the decomposition of matrices
       Per-Gunnar Martinsson, Vladimir Rokhlin and Mark Tygert
     """
+    if n_iterations is not None:
+        warnings.warn("n_iterations was renamed to n_iter for consistency "
+                "and will be removed in 0.16.", DeprecationWarning)
+        n_iter = n_iterations
+
     random_state = check_random_state(random_state)
     n_random = n_components + n_oversamples
     n_samples, n_features = M.shape
@@ -182,7 +193,7 @@ def randomized_svd(M, n_components, n_oversamples=10, n_iterations=0,
         # this implementation is a bit faster with smaller shape[1]
         M = M.T
 
-    Q = randomized_range_finder(M, n_random, n_iterations, random_state)
+    Q = randomized_range_finder(M, n_random, n_iter, random_state)
 
     # project M to the (k + p) dimensional space using the basis vectors
     B = safe_sparse_dot(Q.T, M)

sklearn/utils/tests/test_svd.py

@@ -69,14 +69,14 @@ def test_randomized_svd_low_rank_with_noise():
 
     # compute the singular values of X using the fast approximate method
     # without the iterated power method
-    _, sa, _ = randomized_svd(X, k, n_iterations=0)
+    _, sa, _ = randomized_svd(X, k, n_iter=0)
 
     # the approximation does not tolerate the noise:
     assert_greater(np.abs(s[:k] - sa).max(), 0.05)
 
     # compute the singular values of X using the fast approximate method with
     # iterated power method
-    _, sap, _ = randomized_svd(X, k, n_iterations=5)
+    _, sap, _ = randomized_svd(X, k, n_iter=5)
 
     # the iterated power method is helping getting rid of the noise:
     assert_almost_equal(s[:k], sap, decimal=3)
@@ -100,14 +100,14 @@ def test_randomized_svd_infinite_rank():
 
     # compute the singular values of X using the fast approximate method
     # without the iterated power method
-    _, sa, _ = randomized_svd(X, k, n_iterations=0)
+    _, sa, _ = randomized_svd(X, k, n_iter=0)
 
     # the approximation does not tolerate the noise:
     assert_greater(np.abs(s[:k] - sa).max(), 0.1)
 
     # compute the singular values of X using the fast approximate method with
     # iterated power method
-    _, sap, _ = randomized_svd(X, k, n_iterations=5)
+    _, sap, _ = randomized_svd(X, k, n_iter=5)
 
     # the iterated power method is still managing to get most of the structure
     # at the requested rank
@@ -125,11 +125,11 @@ def test_randomized_svd_transpose_consistency():
         effective_rank=rank, tail_strength=0.5, random_state=0)
     assert_equal(X.shape, (n_samples, n_features))
 
-    U1, s1, V1 = randomized_svd(X, k, n_iterations=3, transpose=False,
+    U1, s1, V1 = randomized_svd(X, k, n_iter=3, transpose=False,
                                 random_state=0)
-    U2, s2, V2 = randomized_svd(X, k, n_iterations=3, transpose=True,
+    U2, s2, V2 = randomized_svd(X, k, n_iter=3, transpose=True,
                                 random_state=0)
-    U3, s3, V3 = randomized_svd(X, k, n_iterations=3, transpose='auto',
+    U3, s3, V3 = randomized_svd(X, k, n_iter=3, transpose='auto',
                                 random_state=0)
     U4, s4, V4 = linalg.svd(X, full_matrices=False)