第二周

模型评判:

对测试数据探索

import numpy as np

from sklearn import datasets

import matplotlib.pyplot as plt

iris = datasets.load_iris()

X = iris.data

y = iris.target

X.shape

Out[3]:

(150, 4)

In [4]:

# 方法1# 使用concatenate函数进行拼接,因为传入的矩阵必须具有相同的形状。因此需要对label进行reshape操作,reshape(-1,1)表示行数自动计算,1列。axis=1表示纵向拼接。

tempConcat = np.concatenate((X, y.reshape(-1,1)), axis=1)

# 拼接好后,直接进行乱序操作np.random.shuffle(tempConcat)

# 再将shuffle后的数组使用split方法拆分

shuffle_X,shuffle_y = np.split(tempConcat, [4], axis=1)

# 设置划分的比例

test_ratio = 0.2

test_size = int(len(X) * test_ratio)

X_train = shuffle_X[test_size:]

y_train = shuffle_y[test_size:]

X_test = shuffle_X[:test_size]

y_test = shuffle_y[:test_size]

In [6]:

print(X_train.shape)

print(X_test.shape)

print(y_train.shape)

print(y_test.shape)

(120, 4)

(30, 4)

(120, 1)

(30, 1)

In [8]:

# 方法2# 将x长度这么多的数,返回一个新的打乱顺序的数组,注意,数组中的元素不是原来的数据,而是混乱的索引

shuffle_index = np.random.permutation(len(X))

# 指定测试数据的比例

test_ratio = 0.2

test_size = int(len(X) * test_ratio)

test_index = shuffle_index[:test_size]

train_index = shuffle_index[test_size:]

X_train = X[train_index]

X_test = X[test_index]

y_train = y[train_index]

y_test = y[test_index]

In [9]:

print(X_train.shape)

print(X_test.shape)

print(y_train.shape)

print(y_test.shape)

(120, 4)

(30, 4)

(120,)

(30,)

In [17]:

import numpy as np

def train_test_split(X, y, test_ratio=0.2, seed=None):

    assert X.shape[0] == y.shape[0]

    assert 0.0 <= test_ratio <= 1.0     

    if seed:    

        # 是否使用随机种子,使随机结果相同,方便debug

        np.random.seed(seed)    

        # permutation(n) 可直接生成一个随机排列的数组,含有n个元素

    shuffle_index = np.random.permutation(len(X))

    test_size = int(len(X) * test_ratio)

    test_index = shuffle_index[:test_size]

    train_index = shuffle_index[test_size:]

    X_train = X[train_index]

    X_test = X[test_index]

    y_train = y[train_index]

    y_test = y[test_index]    

    return X_train, X_test, y_train, y_test

In [21]:

from myAlgorithm.kNN import kNNClassifier

my_kNNClassifier = kNNClassifier(k=3)

my_kNNClassifier.fit(X_train, y_train)

y_predict = my_kNNClassifier.predict(X_test)

y_predict

y_test

# 两个向量的比较,返回一个布尔型向量,对这个布尔向量(faluse=1,true=0)sum,

sum(y_predict == y_test)

sum(y_predict == y_test)/len(y_test)

---------------------------------------------------------------------------

ModuleNotFoundError                       Traceback (most recent call last)

in

----> 1 from myAlgorithm.kNN import kNNClassifier

      2 my_kNNClassifier = kNNClassifier(k=3)

      3 my_kNNClassifier.fit(X_train, y_train)

      4 y_predict = my_kNNClassifier.predict(X_test)

      5 y_predict

ModuleNotFoundError: No module named 'myAlgorithm'

In [16]:

from sklearn.model_selection import train_test_split

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=666)

print(X_train.shape)

print(X_test.shape)

print(y_train.shape)

print(y_test.shape)

(120, 4)

(30, 4)

(120,)

(30,)

import numpy as np

import matplotlib

import matplotlib.pyplot as plt

from sklearn import datasets

from sklearn.model_selection

import train_test_split

from sklearn.neighbors import KNeighborsClassifier

# 手写数字数据集,封装好的对象,可以理解为一个字段

digits = datasets.load_digits()

# 可以使用keys()方法来看一下数据集的详情digits.keys()

# 输出:dict_keys(['data', 'target', 'target_names', 'images', 'DESC

# 特征的shapeX = digits.data

X.shape

(1797, 64)

# 标签的shape

y = digits.target

y.shape

(1797, )

# 标签分类digits.target_names

array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9])

# 去除某一个具体的数据,查看其特征以及标签信息some_digit = X[666]

some_digit

array([ 0.,  0.,  5., 15., 14.,  3.,  0.,  0.,  0.,  0., 13., 15.,  9.,15.,  2.,  0.,  0.,  4., 16., 12.,  0., 10.,  6.,  0.,  0.,  8.,16.,  9.,  0.,  8., 10.,  0.,  0.,  7., 15.,  5.,  0., 12., 11.,0.,  0.,  7., 13.,  0.,  5., 16.,  6.,  0.,  0.,  0., 16., 12.,15., 13.,  1.,  0.,  0.,  0.,  6., 16., 12.,  2.,  0.,  0.])

y[666]

# 也可以这条数据进行可视化

some_digmit_image = some_digit.reshape(8, 8)

plt.imshow(some_digmit_image, cmap = matplotlib.cm.binary)

plt.show()

# 指定最佳值的分数,初始化为0.0;设置最佳值k,初始值为-1

best_score = 0.0

best_k = -1

for k in range(1, 11):  # 暂且设定到1~11的范围内

    knn_clf = KNeighborsClassifier(n_neighbors=k)

    knn_clf.fit(X_train, y_train)

    score = knn_clf.score(X_test, y_test)    

    if score > best_score:

        best_k = k

        best_score = score

print("best_k = ", best_k)

print("best_score = ", best_score)

# 输出:best_k =  4best_score =  0.9916666666666667

# 两种方式进行比较

best_method = ""

best_score = 0.0

best_k = -1

for method in ["uniform","distance"]:    

    for k in range(1, 11):

        knn_clf = KNeighborsClassifier(n_neighbors=k, weights=method, p=2)

        knn_clf.fit(X_train, y_train)

        score = knn_clf.score(X_test, y_test)        

        if score > best_score:

            best_k = k

            best_score = score

            best_method = method

print("best_method = ", method)

print("best_k = ", best_k)

print("best_score = ", best_score)

# 输出:best_method =  distance

# best_k =  4best_score =  0.9916666666666667

你可能感兴趣的:(第二周)