Easyensemble&LightGBM-应对气象样本不平衡问题的有效算法（支持各类基模型接入与新增优化参数）

#%%
import pandas as pd
from Ensemble_model import easyensemble
import warnings
import os
import numpy as np 
import matplotlib.pyplot as plt
from collections import Counter
from sklearn.datasets import make_classification
import meteva.method as mem
import pickle
warnings.filterwarnings("ignore")
#%%
# 1.创建类别不平衡的数据集：
# 使用make_classification生成样本数据
X, Y = make_classification(n_samples=5000,
                           n_features=2,  # 特征个数 = n_informative() + n_redundant + n_repeated
                           n_informative=2,  # 多信息特征的个数
                           n_redundant=0,  # 冗余信息，informative特征的随机线性组合
                           n_repeated=0,  # 重复信息，随机提取n_informative和n_redundant 特征
                           n_classes=2,  # 分类类别
                           n_clusters_per_class=1,  # 某一个类别是有几个cluster构成的
                           weights=[0.99, 0.01],  # 列表类型，权重比
                           random_state=0
                           )

# 2.查看各个标签的样本：
counter = Counter(Y)

# 3.数据集可视化：
plt.scatter(X[:, 0], X[:, 1], c=Y)
plt.show()

path_model_save = 'Data/model/'  # 将easyensemble训练得到的模型保存，以便后续调用
path_importance_save = 'Data/importance/' # 获得单个基模型对应的特征重要性

cols_x = [str(i) for i in range(X.shape[1])]
num = 10
rate = 3

from sklearn.model_selection import train_test_split
xtrain,xvalid,ytrain,yvalid = train_test_split(X,Y,test_size=0.6,random_state=2021)
xvalid,xtest,yvalid,ytest = train_test_split(xvalid,yvalid,test_size=0.5,random_state=2021)

easyensemble(xtrain,
            ytrain,
            xvalid,
            yvalid,
            save_path=path_model_save,
            importance_save_path=path_importance_save,
            features_name = cols_x,
            rate = rate,num = num,weight = [1, 1])

###调用已经保存的模型
mds = os.listdir(path_model_save)
test_prob = np.empty((xtest.shape[0], len(mds)))  # 测试集
valid_prob = np.empty((xvalid.shape[0], len(mds)))  # 测试集
for idx,md in enumerate(mds):
    pic2 = open(path_model_save + md,'rb')
    md_load = pickle.load(pic2)
    test_prob[:, idx] = md_load.predict_proba(xtest)[:, -1]
    valid_prob[:, idx] = md_load.predict_proba(xvalid)[:, -1]
test_prob_result = np.average(test_prob, axis=1)
valid_prob_result = np.average(valid_prob, axis=1)

predtest_easyensemble = test_prob_result.copy()
predtest_easyensemble[test_prob_result >= 0.5] = 1
predtest_easyensemble[test_prob_result < 0.5] = 0

predvalid_easyensemble = valid_prob_result.copy()
predvalid_easyensemble[valid_prob_result >= 0.5] = 1
predvalid_easyensemble[valid_prob_result < 0.5] = 0

def mkdir(path):
    # 引入模块
    import os
    # 去除首位空格
    path = path.strip()
    # 去除尾部 \ 符号
    path = path.rstrip("\\")
    # 判断路径是否存在
    # 存在     True
    # 不存在   False
    isExists = os.path.exists(path)
    # 判断结果
    if not isExists:
        # 如果不存在则创建目录
        # 创建目录操作函数
        os.makedirs(path)
        # print(path + ' 创建成功')
        return True
    else:
        # 如果目录存在则不创建，并提示目录已存在
        # print(path + ' 目录已存在')
        return False