knn 数据集准备

 """
 Created on Mon Aug 26 20:57:24 2019

 @author: huoqs

 knn algorithm
 """
 import numpy as np
 import matplotlib.pyplot as plt

 def generate_data(num_samples, num_features=2):
     data_size = (num_samples, num_features)
     data = np.random.randint(0, 100, data_size)

     label_size = (num_samples, 1)
     labels = np.random.randint(0, 2, label_size)
     # must be float32
     return data.astype(np.float32), labels

 def plot_data(all_blue, all_red):
     plt.scatter(all_blue[:, 0], all_blue[:, 1], c = 'b', marker = 's', s = 180)
     plt.scatter(all_red[:, 0], all_red[:, 1], c = 'r', marker = '^', s = 180)
     plt.xlabel('x')
     plt.ylabel('y')

 plt.style.use('ggplot')

 np.random.seed(42)

 train_data, labels = generate_data(11)

 # print(train_data, labels)

 blue = train_data[labels.ravel() == 0]
 red = train_data[labels.ravel() == 1]

 plot_data(blue, red)

知识点：

1、np.random.randint 函数，生成一个数组，参数：low,high,size,type

https://docs.scipy.org/doc/numpy-1.15.1/reference/generated/numpy.random.randint.html

2、ndarray.ravel()，将数组扁平化，变为一维数组

https://docs.scipy.org/doc/numpy/reference/generated/numpy.ndarray.ravel.html