mean_shift.py

import numpy as np
import matplotlib.pyplot as plt
from matplotlib import style
style.use('ggplot')

fig, ax  = plt.subplots()

norm = lambda x1, x2: np.linalg.norm(x1-x2)
def draw_circle(x=0, y=0, r=1):
    circle = plt.Circle( (x, y), r, alpha=0.35)
    ax.add_artist(circle)
def neighbors(point, X, distance):
    friends = [] 
    for x in X:
        if( norm(point, x) <= distance):
            friends.append(x)
    return friends
def kernel(distance, bandwidth):
    return (1 / (bandwidth * np.sqrt(2 * np.pi))) * np.exp(-0.5 * ((distance / bandwidth)) ** 2)


xs = np.arange(1, 51) # xs [1-50]
ys = 50 * np.random.random((1, 50))[0] # ys randoms [0-1]
X = np.column_stack((xs, ys)) # X is featureset [xs, ys]


class Mean_Shift:
    def __init__(self, bandwidth=4):
        self.bandwidth = bandwidth
    def fit(self, data):
        centroids = {}
        # make all data points centroids
        for i in range(len(data)):
            centroids[i] = data[i]