yamaguchiyuto

import sys
import random
import os
import tweepy

def get_auth(key_file):
    consumer_key, consumer_secret, access_token, access_token_secret = open(key_file, 'r').readline().rstrip().split(' ')
    auth = tweepy.OAuthHandler(consumer_key, consumer_secret)
    auth.set_access_token(access_token, access_token_secret)
    return auth

yamaguchiyuto

import sys
import networkx as nx
from scipy.sparse import linalg
import matplotlib
matplotlib.use('Agg')
import matplotlib.pyplot as plt

def plot(v,name,k,n=10000):
    for i in range(0,k-1):
        plt.plot(v[:n,i],v[:n,i+1],'r+')

yamaguchiyuto

import matplotlib.pyplot as plt
from sklearn.datasets import load_digits
digits = load_digits()
plt.gray()
plt.matshow(digits.images[0])
plt.show()

yamaguchiyuto

import sys
import numpy as np
import random

from sklearn import datasets
from sklearn.semi_supervised import label_propagation
from sklearn import svm
from sklearn.grid_search import ParameterGrid

def score(estimator, X, y, parameters, validation_true_labels, test_true_labels, validation_set, test_set, X_validation_for_svm=None, X_test_for_svm=None):

yamaguchiyuto

import sys
import numpy as np
from sklearn import cross_validation
from oreore_ridge import RidgeRegression

def psi(xlist,M):
    """ make a design matrix """
    ret = []
    for x in xlist:
        ret.append([x**i for i in range(0,M+1)])

yamaguchiyuto

import numpy as np
from sklearn.base import BaseEstimator, RegressorMixin

class RidgeRegression(BaseEstimator, RegressorMixin):
    def __init__(self,lamb=1.0):
        self.lamb = lamb

    def fit(self,X,y):
        A = np.dot(X.T,X) + self.lamb * np.identity(X.shape[1])
        b = np.dot(X.T,y)

yamaguchiyuto

import sys
import numpy as np
import matplotlib.pyplot as plt
from sklearn import grid_search
from oreore_ridge import RidgeRegression

def psi(xlist,M):
    """ make a design matrix """
    ret = []
    for x in xlist:

yamaguchiyuto

import re

def remove_usernames_and_urls(text):
    username_removed_text = re.sub('@\w+', '', text) # remove usernames
    return re.sub('(https?|ftp)(:\/\/[-_.!~*\'()a-zA-Z0-9;\/?:\@&=+\$,%#]+)', '', username_removed_text) # remove urls

yamaguchiyuto

import time
 
twitter_timestamp_str = "Tue Apr 16 04:00:29 +0000 2013"
 
format_str = "%a %b %d %H:%M:%S +0000 %Y"
 
encoded_timestamp = time.strptime(twitter_timestamp_str, format_str)
 
print time.mktime(encoded_timestamp)

yamaguchiyuto

import numpy
from scipy.spatial import distance
from sklearn.cluster import DBSCAN

S = numpy.array([[0,0.9],[0.1,0.8],[0.9,0.1],[0.85,0.05],[0.9,0.05],[0.05,0.85],[0.5,0.4]])
dbs = DBSCAN(eps=0.2, min_samples=3)
dbs.fit(S)
dbs.labels_  # => array([ 1.,  1.,  0.,  0.,  0.,  1., -1.])