mblondel’s gists

## local_regression.py
"""Local regression"""

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

import numpy as np

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

## echo_server.py
# adapted from http://roscidus.com/desktop/node/413

import socket
import gobject

def server(host, port):
    '''Initialize server and start listening.'''
    sock = socket.socket()
    sock.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR, 1)
    sock.bind((host, port))

## coroutines.py
def recv_count():
    try:
        while True:
             n = (yield)
             print "T-minus", n
    except GeneratorExit:
        print "Kaboom!"

def ex1():
    r = recv_count()

## number_plate_solver.py
#!/usr/bin/env python

"""
Find the operations needed to sum up to TARGET by using all 4 numbers in NUMBERS.
"""

from itertools import permutations, product

NUMBERS = ["3","4","7","8"]
TARGET = 10.0

## mc_pi.py
from random import random

"""
Find pi by the Monte-Carlo method.

area of a circle = pi r^2
area of a square = (2r)^2 = 4 r^2

Perform random uniform sampling between -1 and 1.
The proportion of points in the unit circle is:

## regression_lp.py
# (C) 2011 Mathieu Blondel
# License: BSD 3 clause

import numpy as np
import numpy.linalg as linalg
import pylab as pl
from cvxopt import matrix, solvers

np.random.seed(0)

## online_variance.py

def online_mean_variance(iterable):
    mN = 0
    mM = 0.0
    mS = 0.0

    for x in iterable:
        mN += 1
        nextM = mM + (x - mM) / mN
        mS += (x - mM) * (x - nextM)

## lbfgs_nnls.py
# (C) Mathieu Blondel 2012
# License: BSD 3 clause

import numpy as np
from scipy.optimize import fmin_l_bfgs_b

from sklearn.base import BaseEstimator, RegressorMixin
from sklearn.utils.extmath import safe_sparse_dot


## xref.txt

paper.tex: main manuscript
supp.tex: supplementary material

Cross-referencing
-----------------

We want to cross-reference equations in paper.tex from supp.tex.


## sparse_multiclass_numba.py
"""
(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.