感知器演算法及其python 實現 V2.0

import matplotlib.pyplot as plt

import numpy as np

from random import uniform, seed, shuffle ,sample

import math

import logging

# from random import random

'''JY_Toolkit.py'''

class Jy_makeDataset(object):

     def random_state(random_seed):

         seed(int(random_seed))

     def draw_HalfMoon(n_sample: int = 1000,       # 樣本點個數，兩個分類一共 n_sample

                       w: float = 1,              # 半月的線寬

                       radius: float = 4,         # 半月的半徑

                       hor_distance: float = 4,   # Horizontal direction distance for two point

                       ver_distance: float = 0,   # Vertical direction distance for two point

                       slope: float = 0,          # 半月傾斜的角度   [0 ~ 180]

                       positive_val: int = 1,

                       negative_val: int = -1,

                       ):

         slope %= 180            # make the `slope`  between 0 and 180

         # 將 n_sample 和樣本分為兩類每個樣本 n_sample / 2 類

         each_m = n_sample//2

         # circle origin point of positive moon [x , y]

         p_origin = [1 + w/2 + radius, 1 + w/2 + radius + ver_distance]

         # circle origin point of negative moon [x , y]

         n_origin = [p_origin[0] + hor_distance, p_origin[1] - ver_distance]

         # product positive point

         p_sample = []

         n_sample = []

         for i in range(each_m):

             # Randomly generate l

             temp_l = radius + uniform(-(w/2), w/2)

             # Randomly generate angle i.e. theta

             temp_angle = uniform(slope, slope + 180)

             point_x = p_origin[0] + temp_l*math.cos(math.pi/180*temp_angle)

             point_y = p_origin[1] + temp_l*math.sin(math.pi/180*temp_angle)

             p_sample.append([point_x, point_y, positive_val])

         for i in range(each_m):

             # Randomly generate l

             temp_l = radius + uniform(-(w/2), w/2)

             # Randomly generate angle i.e. theta , but the angle of negative point should between `slope + 180` and `slope + 360`

             temp_angle = uniform(slope + 180, slope + 360)

             point_x = n_origin[0] + temp_l*math.cos(math.pi/180*temp_angle)

             point_y = n_origin[1] + temp_l*math.sin(math.pi/180*temp_angle)

             n_sample.append([point_x, point_y, negative_val])

         sample_points = p_sample + n_sample

         shuffle(sample_points)

         sample_points = np.array(sample_points)

         return sample_points[:, 0:2], sample_points[:, 2]

     pass

class Jy_dataSetProcess(object):

     def Jy_train_test_split(X,

                             y,

                             test_size : 0.2,

                             ):

         data = np.column_stack((X,y))

         if test_size >= 1 and test_size <= 0:

             logging.exception('test_size must be greater than 0 less than 1, we will assign test_size value of 0.2')

             test_size = 0.2

         sample_count = int(len(data)*test_size)

         '''

         分離思路：

         先將輸入的資料集打亂，外匯跟單gendan5.com然後取前 test_size 部分為測試集，後部分為訓練集

         '''

         shuffle(data)

         X_test = data[0:sample_count-1]

         X_train = data[sample_count:]

         return X_train[:,0:2],  X_test[:,0:2] ,X_train[:,2] , X_test[:,2]

     pass

if __name__ == '__main__':

     random_seed = 52

     Jy_makeDataset.random_state(random_seed)

     np_data, label = Jy_makeDataset.draw_HalfMoon(n_sample=2000)

     p_point_x1 = [np_data[i][0] for i in range(len(np_data)) if label[i] == 1]

     p_point_x2 = [np_data[i][1] for i in range(len(np_data)) if label[i] == 1]

     n_point_x1 = [np_data[i][0] for i in range(len(np_data)) if label[i] == -1]

     n_point_x2 = [np_data[i][1] for i in range(len(np_data)) if label[i] == -1]

     fig = plt.figure(num="HalfMoons", figsize=(8, 8))

     ax1 = fig.add_subplot(111)

     ax1.scatter(p_point_x1, p_point_x2, c='red')

     ax1.scatter(n_point_x1, n_point_x2, c='blue')

     plt.show()

     print(np_data)