import numpy as npimport pandas as pdfrom pandas import DataFrame# 读取数据train= pd.read_csv('./data/train_V2.csv')# 查看列索引，为后续选择特征做准备train.columns# 选择特征data = train[['assists', 'DBNOs', 'headshotKills', 'heals', 'killPlace', 'killPoints', 'kills', 'killStreaks', 'matchDuration', 'maxPlace', 'rankPoints', 'revives', 'roadKills','weaponsAcquired', 'winPoints','winPlacePerc']]# 查看选择的数据data# 处理数据中的异常值data = data.dropna()# 查看处理数据异常值后的数据data#检查是否存在异常值np.all(pd.notnull(data))np.all(pd.notna(data))# 选择标签值target = data['winPlacePerc']#检查是否存在异常值np.all(pd.notnull(target))np.all(pd.notna(target))# 查看选择的数据target# 开始进入模型的编写from sklearn.model_selection import train_test_splitfrom sklearn.preprocessing import StandardScalerfrom sklearn.linear_model import LinearRegressionfrom sklearn.metrics import mean_squared_error# 定义模型函数def linear_model():#分割数据集x_train,x_test,y_train,y_test = train_test_split(data,target,test_size=0.2)#标准化数据transfer = StandardScaler()x_train = transfer.fit_transform(x_train)x_test = transfer.fit_transform(x_test)# 设置一个 模型estimator = LinearRegression()# 模型训练estimator.fit(x_train,y_train)# 模型预测y_predict = estimator.predict(x_test)print(y_predict)# 模型分数score = estimator.score(x_test,y_test)print('score',score)# 模型均方误差ret = mean_squared_error(y_test,y_predict)print(ret)linear_model()
输出结果如下
[1.00096424 0.17929419 0.50079986 ... 0.07756076 0.1929987 0.6965642 ]score 0.99999327205917346.354406278810655e-07
