e2epp/build_predict_model.py at master · yn-sun/e2epp · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
import numpy as np
from sklearn.tree import DecisionTreeRegressor
import pandas as pd
import matplotlib.pyplot as plt

def load_data():

    d = pd.read_csv('data.txt').values
    data = d[:, 1:-1]
    label = d[:,-1]
    return data, label

def make_decision_trees(train_data, train_label, n_tree):
    feature_record = []
    tree_record = []
    for _ in range(n_tree):
        sample_idx = np.arange(train_data.shape[0])
        np.random.shuffle(sample_idx)
        train_data = train_data[sample_idx, :]
        train_label = train_label[sample_idx]

        feature_idx = np.arange(train_data.shape[1])
        np.random.shuffle(feature_idx)
        n_feature = np.random.randint(1, train_data.shape[1]+1)
        selected_feature_ids = feature_idx[0:n_feature]
        feature_record.append(selected_feature_ids)

        dt = DecisionTreeRegressor()
        dt.fit(train_data[:, selected_feature_ids], train_label)
        tree_record.append(dt)
    return tree_record, feature_record

def predict(test_data, trees, feature_ids):
    predict_list = []
    for tree, feature in zip(trees, feature_ids):
        predict_y = tree.predict(test_data[:, feature])
        predict_list.append(predict_y)
    return predict_list

def one_run(train_data, train_label, test_data, test_label):
    n_tree = 1000
    trees, features = make_decision_trees(train_data, train_label, n_tree)
    total_predict_list = np.zeros((len(trees), test_label.shape[0]))
    for i, (tree, feature) in enumerate(zip(trees, features)):
        predict_list = tree.predict(test_data[:, feature])
        total_predict_list[i,:] = predict_list
    predict_mean_y = np.mean(total_predict_list, 0)
    diff = np.mean(np.square(predict_mean_y - test_label))
    return diff, predict_mean_y, test_label


def one_run_for_each(train_data, train_label, test_data, test_label):
    n_tree = 1000
    trees, features = make_decision_trees(train_data, train_label, n_tree)
    test_num = test_data.shape[0]
    predict_labels = np.zeros(test_num)
    for i in range(test_num):
        this_test_data = test_data[i,:]
        predict_this_list = np.zeros(n_tree)

        for j, (tree, feature) in enumerate(zip(trees, features)):
            predict_this_list[j] =tree.predict([this_test_data[feature]])[0]

        # find the top 100 prediction
        predict_this_list = np.sort(predict_this_list)
        predict_this_list = predict_this_list[::-1]
        this_predict = np.mean(predict_this_list[0:100])
        predict_labels[i] = this_predict
    print(np.sqrt(np.mean(np.square(predict_labels - test_label))))
    print(np.mean(np.abs(predict_labels - test_label)))

    plt.plot(predict_labels, label='predict')
    plt.plot(test_label, label='true')
    plt.legend()
    plt.show()

def train_final_model():
    n_tree = 5000
    train_data, train_label = load_data()
    trees, features = make_decision_trees(train_data, train_label, n_tree)
    model = [trees, features]
    from sklearn.externals import joblib
    joblib.dump(model, 'predict_model.pkl')

def test_saved_model():
    n_tree = 5000
    from sklearn.externals import joblib
    test_data, test_label = load_data()
    test_num = test_data.shape[0]
    trees, features = joblib.load('predict_model.pkl')
    print(type(trees), len(trees))

    predict_labels = np.zeros(test_num)
    for i in range(test_num):
        this_test_data = test_data[i,:]
        predict_this_list = np.zeros(n_tree)

        for j, (tree, feature) in enumerate(zip(trees, features)):
            predict_this_list[j] =tree.predict([this_test_data[feature]])[0]

        # find the top 100 prediction
        predict_this_list = np.sort(predict_this_list)
        predict_this_list = predict_this_list[::-1]
        this_predict = np.mean(predict_this_list)
        predict_labels[i] = this_predict
    print(np.sqrt(np.mean(np.square(predict_labels - test_label))))
    print(np.mean(np.abs(predict_labels - test_label)))

    plt.plot(predict_labels, label='predict')
    plt.plot(test_label, label='true')
    plt.legend()
    plt.show()

def test_one_this_run():
    data, label = load_data()
    idx = np.arange(label.shape[0])
#     np.random.shuffle(idx)
#     data = data[idx,:]
#     label = label[idx]
    train_num = int(idx.shape[0]*0.8)
    train_data = data[0:train_num, :]
    train_label = label[0:train_num]
    test_data = data[train_num:, :]
    test_label = label[train_num:]
    err = one_run_for_each(train_data, train_label, test_data, test_label)
    print(err)


def ten_folds():
    folds = 10
    data, label = load_data()
    fold_size = 10 #data.shape[0]//folds
    fold_mean = np.zeros(folds)
    predict_y_list = []
    true_y_list = []
    for i in range(folds):
        start_index = i*fold_size
        end_index = (i+1)*fold_size
        copy_data = np.copy(data)
        copy_label = np.copy(label)
        test_data = copy_data[start_index:end_index, :]
        test_label = copy_label[start_index:end_index]
        train_data = np.delete(copy_data, np.arange(start_index, end_index), axis=0)
        train_label = np.delete(copy_label, np.arange(start_index, end_index))
        diff, predict_y, true_y = one_run(train_data, train_label, test_data, test_label)
        fold_mean[i] = diff
        predict_y_list.append(predict_y)
        true_y_list.append(true_y)
    print('MSE:', np.mean(fold_mean))


if __name__ == '__main__':
    ten_folds()