scikit-network 0.30.0__cp39-cp39-win_amd64.whl → 0.32.1__cp39-cp39-win_amd64.whl

sknetwork/linalg/diteration.pyx

@@ -1,7 +1,5 @@
 # distutils: language = c++
 # cython: language_level=3
-# cython: linetrace=True
-# distutils: define_macros=CYTHON_TRACE_NOGIL=1
 """
 Created on Apr 2020
 @author: Nathan de Lara <nathan.delara@polytechnique.org>

sknetwork/linalg/eig_solver.py

@@ -10,7 +10,7 @@ from typing import Union
 import numpy as np
 from scipy import sparse
 from scipy.sparse.linalg import eigsh
-from sknetwork.utils.base import Algorithm
+from sknetwork.base import Algorithm
 
 
 class EigSolver(Algorithm, ABC):

sknetwork/linalg/{normalization.py → normalizer.py}

@@ -35,30 +35,33 @@ def get_norms(matrix: Union[sparse.csr_matrix, np.ndarray, LinearOperator], p=1)
 
     Parameters
     ----------
-    matrix : numpy array, sparse CSR matrix or linear operator, shape (n_rows, n_cols)
+    matrix : numpy array or sparse CSR matrix or LinearOperator, shape (n_rows, n_cols)
         Input matrix.
     p :
-        Order of the norm.
+        Order of the norm (1 or 2).
     Returns
     -------
     norms : np.array, shape (n_rows,)
         Vector norms
     """
+    n_row, n_col = matrix.shape
+    if isinstance(matrix, np.ndarray):
+        input_matrix = sparse.csr_matrix(matrix)
+    elif isinstance(matrix, sparse.csr_matrix):
+        input_matrix = matrix.copy()
+    else:
+        input_matrix = matrix
     if p == 1:
-        norms = matrix.dot(np.ones(matrix.shape[1]))
+        if not isinstance(matrix, LinearOperator):
+            input_matrix.data = np.abs(input_matrix.data)
+        return input_matrix.dot(np.ones(n_col))
     elif p == 2:
-        if isinstance(matrix, np.ndarray):
-            norms = np.linalg.norm(matrix, axis=1)
-        elif isinstance(matrix, sparse.csr_matrix):
-            data = matrix.data.copy()
-            matrix.data = data ** 2
-            norms = np.sqrt(matrix.dot(np.ones(matrix.shape[1])))
-            matrix.data = data
-        else:
-            raise NotImplementedError('Norm 2 is not available for a LinearOperator.')
+        if isinstance(matrix, LinearOperator):
+            raise ValueError('Only norm 1 is available for linear operators.')
+        input_matrix.data = input_matrix.data**2
+        return np.sqrt(input_matrix.dot(np.ones(n_col)))
     else:
-        raise NotImplementedError('Only norms 1 and 2 are available at the moment.')
-    return norms
+        raise ValueError('Only norms 1 and 2 are available.')
 
 
 def normalize(matrix: Union[sparse.csr_matrix, np.ndarray, LinearOperator], p=1):
@@ -74,7 +77,7 @@ def normalize(matrix: Union[sparse.csr_matrix, np.ndarray, LinearOperator], p=1)
     Returns
     -------
     normalized matrix :
-        Normalized matrix.
+        Normalized matrix (same format as input matrix).
     """
     norms = get_norms(matrix, p)
     diag = diagonal_pseudo_inverse(norms)

sknetwork/linalg/operators.py

@@ -12,7 +12,7 @@ from scipy import sparse
 from scipy.sparse.linalg import LinearOperator
 
 from sknetwork.linalg import diagonal_pseudo_inverse
-from sknetwork.linalg.normalization import normalize
+from sknetwork.linalg.normalizer import normalize
 from sknetwork.linalg.sparse_lowrank import SparseLR
 from sknetwork.utils.check import check_format
 

sknetwork/linalg/ppr_solver.py

@@ -12,7 +12,7 @@ from scipy.sparse.linalg import eigs, LinearOperator, bicgstab
 
 from sknetwork.linalg.diteration import diffusion
 from sknetwork.linalg.push import push_pagerank
-from sknetwork.linalg.normalization import normalize
+from sknetwork.linalg.normalizer import normalize
 from sknetwork.linalg.polynome import Polynome