mblondel

# -*- coding: utf-8 -*-

# Copyright (C) 2010 Mathieu Blondel
#
# This program is free software; you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation; either version 2 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,

mblondel

# Authors: Mathieu Blondel, Vlad Niculae
# License: BSD 3 clause

import numpy as np


def _gen_pairs(gen, max_iter, max_inner, random_state, verbose):
    rng = np.random.RandomState(random_state)

    # if tuple, interpret as randn

mblondel

"""
(C) August 2013, Mathieu Blondel
# License: BSD 3 clause

This is a Numba-based reimplementation of the block coordinate descent solver
(without line search) described in the paper:

    Block Coordinate Descent Algorithms for Large-scale Sparse Multiclass
    Classification.  Mathieu Blondel, Kazuhiro Seki, and Kuniaki Uehara.
    Machine Learning, May 2013.

mblondel

# Author: Mathieu Blondel
# License: BSD 3 clause

import numpy as np
from sklearn.isotonic import isotonic_regression


def projection_simplex(x, z=1):
    """
    Compute argmin_{p : p >= 0 and \sum_i p_i = z} ||p - x||

mblondel

"""
NMF by coordinate descent, designed for sparse data (without missing values)
"""

# Author: Mathieu Blondel <mathieu@mblondel.org>
# License: BSD 3 clause

import numpy as np
import scipy.sparse as sp
import numba

mblondel

# (C) Mathieu Blondel, November 2013
# License: BSD 3 clause

import numpy as np
from scipy.linalg import svd


def frequent_directions(A, ell, verbose=False):
    """
    Return the sketch of matrix A.

mblondel

"""Gaussian processes"""

# Author: Mathieu Blondel <mathieu@mblondel.org>
# License: BSD 3 clause

import numpy as np
from scipy.linalg import cholesky, solve_triangular

from sklearn.base import BaseEstimator, RegressorMixin
from sklearn.metrics.pairwise import pairwise_kernels

mblondel

import numpy as np

def _logsum(logx, logy):
    """
    Return log(x+y), avoiding arithmetic underflow/overflow.

    logx: log(x)
    logy: log(y)

    Rationale:

mblondel

def test():
    print(i)

i = 1

test()

mblondel

import numpy as np

rng = np.random.RandomState(0)

print "Trace"
A = rng.rand(3, 3)
print np.trace(A)
print np.einsum("ii", A)
print