【数据分析】图书馆数据-08决策树

# -*-coding:utf-8-*-

import numpy as np
import pandas as pd

"""
    数据要求：read_num, book_id
"""
pf = pd.read_csv('new_data.csv', encoding='gbk')
print type(pf)

unit = pf['read_unit']
unit = unit.str.split(' ')  # 原来是空格分隔的
dapartment = unit.str[0]  # 学院
# print dapartment
major = unit.str[1]  # 专业
# print major

data = pf[['read_sex', 'book_id']]
print type(data)
data.insert(2, 'dapartment', dapartment)
# data.insert(3, 'major', major)
# data.columns = ['sex', 'book', 'dapartment', 'major']
data.columns = ['sex', 'book', 'dapartment']
print type(data)
print '------------------------------------------------------'
"""
    算法：获取标签
"""


def add_label(s):
    l = []
    m = []

    for i in range(len(s)):
        if i == 0:
            m = []
            l = [1]
        else:
            m.append(s[i - 1])
            if s[i] in m:
                if m.index(s[i]) == 0:
                    l.append(1)
                else:
                    l.append(l[m.index(s[i])])
            else:
                l.append(max(l) + 1)
    return l

sex = data['sex']
print type(sex)
sex = add_label(sex)  # 添加标签
dapartment = data['dapartment']
dapartment = add_label(dapartment)  # 添加标签
book = data['book']
book = add_label(book)  # 添加标签
# major = data['major']
# major = add_label(major)
# print 'sex:', sex[:50]
# print 'department:', dapartment[:50]
# print 'book:', book[:50]
# print 'major:', major[:50]

"""
    不同学院的学生借阅书籍的不同
"""
diff_dep = []
diff_dep.append(sex)
diff_dep.append(dapartment)
m = np.array(diff_dep).T
print m  # data
# print len(m)  # 182508
print m[:10]
n = book  # target
print n[150000:]
"""
    决策树训练数据和预测数据
"""
train_data = m[: 150000]
test_data = m[150000:]
train_target = n[: 150000]
test_target = n[150000:]
# 导入决策树DTC包
from sklearn.tree import DecisionTreeClassifier

# 训练
clf = DecisionTreeClassifier()
# 注意均使用训练数据集和样本类标
clf.fit(train_data, train_target)
print clf

# 预测结果
predict_target = clf.predict(test_data)
print predict_target

# 预测结果与真实结果比对
print predict_target == test_target
print sum(predict_target == test_target)

# 输出准确率 召回率 F值
from sklearn import metrics

print metrics.classification_report(test_target, predict_target)
# print metrics.confusion_matrix(test_target, predict_target)