任务三
构建xgboost和lightgbm模型进行预测。
遇到的问题
- 参数不知道怎么调用
- xgboost的接口和sklearn接口不明白
- LGB和XGB自带接口预测(predict)的都是概率
- 训练之前都要将数据转化为相应模型所需的格式
代码
特征处理
import pickle
import pandas as pd #数据分析
from pandas import Series,DataFrame
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import StandardScaler
import time
print("开始......")
t_start = time.time()
path = "E:/mypython/moxingxuexi/data/"
"""=====================================================================================================================
1 读取数据
"""
print("数据预处理")
data = pd.read_csv(path + 'data.csv',encoding='gbk')
"""=====================================================================================================================
2 数据处理
"""
"""将每一个样本的缺失值的个数作为一个特征"""
temp1=data.isnull()
num=(temp1 == True).astype(bool).sum(axis=1)
is_null=DataFrame(list(zip(num)))
is_null=is_null.rename(columns={0:"is_null_num"})
data = pd.merge(data,is_null,left_index = True, right_index = True, how = 'outer')
"""
1.1 缺失值用100填充
"""
data=DataFrame(data.fillna(100))
"""
1.2 对reg_preference_for_trad 的处理 【映射】
nan=0 境外=1 一线=5 二线=2 三线 =3 其他=4
"""
n=set(data['reg_preference_for_trad'])
dic={}
for i,j in enumerate(n):
dic[j]=i
data['reg_preference_for_trad'] = data['reg_preference_for_trad'].map(dic)
"""
1.2 对source 的处理 【映射】
"""
n=set(data['source'])
dic={}
for i,j in enumerate(n):
dic[j]=i
data['source'] = data['source'].map(dic)
"""
1.3 对bank_card_no 的处理 【映射】
"""
n=set(data['bank_card_no'])
dic={}
for i,j in enumerate(n):
dic[j]=i
data['bank_card_no'] = data['bank_card_no'].map(dic)
"""
1.2 对 id_name的处理 【映射】
"""
n=set(data['id_name'])
dic={}
for i,j in enumerate(n):
dic[j]=i
data['id_name'] = data['id_name'].map(dic)
"""
1.2 对 time 的处理 【删除】
"""
data.drop(["latest_query_time"],axis=1,inplace=True)
data.drop(["loans_latest_time"],axis=1,inplace=True)
data.drop(['trade_no'],axis=1,inplace=True)
status = data.status
# """=====================================================================================================================
# 4 time时间归一化 小时
# """
# data['time'] = pd.to_datetime(data['time'])
# time_now = data['time'].apply(lambda x:int((x-datetime(2018,11,14,0,0,0)).seconds/3600))
# data['time']= time_now
"""=====================================================================================================================
2 划分训练集和验证集，验证集比例为test_size
"""
print("划分训练集和验证集，验证集比例为test_size")
train, test = train_test_split(data, test_size=0.3, random_state=666)
"""
标准化数据
"""
standardScaler = StandardScaler()
train_fit = standardScaler.fit_transform(train)
test_fit = standardScaler.transform(test)
"""=====================================================================================================================
3 分标签和 训练数据
"""
y_train= train.status
train.drop(["status"],axis=1,inplace=True)
y_test= test.status
test.drop(["status"],axis=1,inplace=True)
print("3 保存至本地")
data = (train, test, y_train,y_test)
fp = open( 'E:/mypython/moxingxuexi/feature/V3.pkl', 'wb')
pickle.dump(data, fp)
fp.close()XGB
#!/user/bin/env python
#-*- coding:utf-8 -*-
import pickle
import numpy as np
import pandas as pd
from sklearn.model_selection import train_test_split
from pandas import Series,DataFrame
from xgboost import XGBClassifier
from sklearn.metrics import accuracy_score,f1_score,r2_score
"""
读取特征数据
"""
path= "E:/mypython/moxingxuexi/"
f = open(path + 'feature/V3.pkl','rb')
train,test,y_train,y_test= pickle.load(f)
f.close()
"""
模型训练
"""
print("xgb训练模型")
xgb_model = XGBClassifier()
xgb_model.fit(train,y_train)
"""
模型预测
"""
y_test_pre =xgb_model.predict(test)
"""
模型评估
"""
f1 = f1_score(y_test,y_test_pre,average='macro')
print("f1的分数: {}".format(f1))
r2 = r2_score(y_test,y_test_pre)
print("f2分数：{}".format(r2))
score = xgb_model.score(test,y_test)
print("验证集分数：{}".format(score))
结果
XGB1
用的自带接口与sklearn接口
#!/user/bin/env python
#-*- coding:utf-8 -*-
# @Time : 2018/11/18 15:07
# @Author : 刘
# @Site :
# @File : xgb1.py
# @Software: PyCharm
import pickle
import xgboost as xgb
import pandas as pd
from sklearn.model_selection import train_test_split
from xgboost.sklearn import XGBClassifier
from sklearn import metrics
from sklearn.externals import joblib
print("开始")
"""
读取特征
"""
path = "E:/mypython/moxingxuexi/"
f = open(path+ 'feature/V3.pkl','rb')
train ,test,y_train,y_test=pickle.load(f)
f.close()
"""
将数据格式转换成XGB所需的格式
"""
xgb_val = xgb.DMatrix(test,label= y_test)
xgb_train = xgb.DMatrix(train,y_train)
xgb_test = xgb.DMatrix(test)
"""
模型参数设置
"""
##XGB自带接口
params={
'booster': 'gbtree',#常用的booster有树模型(tree)和线性模型(linear model)
'objective': 'reg:linear',
'gamma': 0.1,#用于控制是否后剪枝的参数，越大越保守一般是0.1,0.2
'max_depth': 10,#构建树的深度，越大越容易过拟合
'lambda': 2,#控制权重值的L2正则化项参数，参数越大，模型越不容易过拟合
'subsample': 0.7,#随机训练样本
'colsample_bytree': 0.7,#生成树时进行的列采样
'min_child_weight': 3,
# 这个参数默认是 1，是每个叶子里面 h 的和至少是多少，对正负样本不均衡时的 0-1 分类而言
#，假设 h 在 0.01 附近，min_child_weight 为 1 意味着叶子节点中最少需要包含 100 个样本。
#这个参数非常影响结果，控制叶子节点中二阶导的和的最小值，该参数值越小，越容易 overfitting。
'silent': 0,#设置成1则没有运行信息输出，最好是设置为0
'eta': 0.001,# 如同学习率
'seed': 1000,#随机种子
# 'nthread':7,# cpu 线程数
#'eval_metric': # 'auc'
}
plst = list(params.items())## 转化为list
num_rounds = 50 # 设置迭代次数
#sklearn接口
##分类使用XGBClassifier
##回归使用XGBRegression
clf = XGBClassifier(
n_estimators =30,
learning_rate =0.3,
max_depth=3,
min_child_weight=1,
gamma=0.3,
subsample=0.8,
colsample_bytree=0.8,
objective= 'binary:logistic',
nthread=12,
scale_pos_weight=1,
reg_lambda=1,
seed=27)
watchlist = [(xgb_train, 'train'),(xgb_val, 'val')]
"""
模型训练
"""
# training model
#early_stopping_rounds 当设置的迭代次数较大时，early_stopping_rounds 可在一定的迭代次数内准确率没有提升就停止训练
# 使用XGBoost有自带接口
"""使用XGBOOST自带训练接口"""
model_xgb= xgb.train(plst,xgb_train,num_rounds,watchlist,early_stopping_rounds=100)
"""使用sklenar接口训练"""
model_xgb_sklearn =clf.fit(train,y_train)
"""
模型保存
"""
print('模型保存')
joblib.dump(model_xgb,path+"model/xgb.pkl")
joblib.dump(model_xgb_sklearn,path+"model/xgb_sklearn.pkl")
"""
模型预测
"""
"""【使用XGB自带接口预测】"""
y_xgb=model_xgb.predict(xgb_test)
"""【使用xgb sklearn预测】"""
y_sklearn_pre= model_xgb_sklearn.predict(test)
y_sklearn_proba= model_xgb_sklearn.predict_proba(test)[:,1]
"""5 模型评分"""
print("XGBoost_自带接口(predict) : %s" % y_xgb)
print("XGBoost_sklearn接口(proba): %s" % y_sklearn_proba)
print("XGBoost_sklearn接口(predict) : %s" % y_sklearn_pre)
# print("XGBoost_自带接口(predict) AUC Score : %f" % metrics.roc_auc_score(y_test, y_xgb))
# print("XGBoost_sklearn接口(proba) AUC Score : %f" % metrics.roc_auc_score(y_test, y_sklearn_proba))
# print("XGBoost_sklearn接口(predict) AUC Score : %f" % metrics.roc_auc_score(y_test, y_sklearn_pre))
"""【roc_auc_score】"""
#直接根据真实值（必须是二值）、预测值（可以是0/1,也可以是proba值）计算出auc值，中间过程的roc计算省略。
# f1 = f1_score(y_test, predictions, average='macro')
print("XGBoost_自带接口(predict) AUC Score ：{}".format(metrics.roc_auc_score(y_test, y_xgb)))
print("XGBoost_sklearn接口(proba) AUC Score : {}".format(metrics.roc_auc_score(y_test, y_sklearn_proba)))
print("XGBoost_sklearn接口(predict) AUC Score ：{}".format(metrics.roc_auc_score(y_test, y_sklearn_pre)))
评分
LGB
#!/user/bin/env python
#-*- coding:utf-8 -*-
# @Time : 2018/11/17 23:03
# @Author : 刘
# @Site :
# @File : lig.py
# @Software: PyCharm
import pickle
import pandas as pd
from pandas import Series,DataFrame
from sklearn.model_selection import train_test_split
from lightgbm import LGBMClassifier
from sklearn.metrics import f1_score,r2_score
"""
读取特征
"""
path= "E:/mypython/moxingxuexi/"
f =open(path + 'feature/V3.pkl','rb')
train,test,y_train,y_test= pickle.load(f)
f.close()
"""
模型训练
"""
print("lgb模型训练")
lgb_model = LGBMClassifier()
lgb_model.fit(train,y_train)
"""
模型预测
"""
y_test_pre = lgb_model.predict(test)
"""
模型评估
"""
f1 = f1_score(y_test,y_test_pre,average='macro')
print("f1的分数: {}".format(f1))
r2 = r2_score(y_test,y_test_pre)
print("f2分数：{}".format(r2))
score = lgb_model.score(test,y_test)
print("验证集分数：{}".format(score))评分
LGB1
#!/user/bin/env python
#-*- coding:utf-8 -*-
# @Time : 2018/11/17 23:03
# @Author : 刘
# @Site :
# @File : lig.py
# @Software: PyCharm
import pickle
import pandas as pd
from pandas import Series,DataFrame
from sklearn.model_selection import train_test_split
from lightgbm import LGBMClassifier
from sklearn.metrics import f1_score,r2_score
"""
读取特征
"""
path= "E:/mypython/moxingxuexi/"
f =open(path + 'feature/V3.pkl','rb')
train,test,y_train,y_test= pickle.load(f)
f.close()
"""
模型训练
"""
print("lgb模型训练")
lgb_model = LGBMClassifier()
lgb_model.fit(train,y_train)
"""
模型预测
"""
y_test_pre = lgb_model.predict(test)
"""
模型评估
"""
f1 = f1_score(y_test,y_test_pre,average='macro')
print("f1的分数: {}".format(f1))
r2 = r2_score(y_test,y_test_pre)
print("f2分数：{}".format(r2))
score = lgb_model.score(test,y_test)
print("验证集分数：{}".format(score))评分
