pearcemc

"""
Quick and dirty experimental implementation of k-nearest neighbours technique
"""

from scipy import randn
import scipy.stats as stats
import random

from kdata import *

pearcemc

from django.conf import settings     #for RECAPTCHA_PRIVATE key

def verify_recaptcha(request):
    import urllib, urllib2, re
    values = {
    "privatekey": settings.RECAPTCHA_PRIVATE,
    "remoteip": request.META['REMOTE_ADDR'],
    "challenge": request.POST['recaptcha_challenge_field'],
    "response": request.POST['recaptcha_response_field'],
    }

pearcemc

def vis3d(ds):
    """Visualise a 3d MultiClassDS"""
    fig = plt.figure()                            #start pyplot
    fig.suptitle('3d vis')                        #add a title
    cols = ['g', 'b', 'r', 'y']                   #colour options
    ax = fig.add_subplot(111, projection="3d")    #initialise 3d
    ax.set_xlabel('X')                            #set axis labels
    ax.set_ylabel('Y')
    ax.set_zlabel('Z')
    for i, clas in enumerate(ds.get_vectors()):   #loop through the vectorised classes 

pearcemc

class MultiClassDS:
    """
    Generates a dataset of n points from c classes in ratio r
    """
    def __init__(self, classes, length=1000, ratio=None):
        #set params
        self.classes, self.length, self.ratio = classes, length, ratio

        #no ratio supplied? set it to even split
        if not ratio: self._normalise_ratio()

pearcemc

import scipy as sp

class PDimClass:
    """
    A p-dimensional class, which takes a list of Normal distribution parameters, 
    each of which defines a dimension
    """
    #a list of normal distribution parameter tuples, for N(mean, sd)
    params = [(0,1), (0,1)]

pearcemc

class TestPredictor:
    """Iterate a KNearest predictor and return its success rate."""

    def __init__(self, predictor, times=1000):
        #set the KNearest predictor
        self.predictor = predictor 
        #set number of iterations   
        self.times = times

    def test(self, k=1):

pearcemc

class KNearest:
    """k-nearest neighbour inferer"""
    def __init__(self, ds):
        #set the dataset
        self.ds = ds

    def predict(self, p1, k=1):
        """Given a test point p1, return the modal class of its knearest neighbours"""
        distances = []
        #calculate the distance between the test point and known data points.

pearcemc

import numpy as np
import matplotlib.pyplot as plt
from math import exp, sqrt, pi

#the function we're going to map
def gausspdf( x, mu, sd2 ):
  return exp( -1*( (x - mu)**2/(2*sd2)) )/sqrt( 2*pi*sd2 )

#canvas size
x = y = np.arange( 0, 10, 0.1 )

pearcemc

##################################################################
#
#       APPLYING EM TO A BINOMIAL MIXTURE MODEL
#       matthew@refute.me.uk 
#       Licence: MIT Licence (http://opensource.org/licenses/MIT)
#
##################################################################


###### EM ALGORITHM FUNCTIONS

pearcemc

import sys
import time
from pylab import *
import scipy.stats as stats

"""
Script to visualise draws from a two parameter Indian Buffet Process.

usage example: