from sklearn.datasets import make_blobs from sklearn.neighbors import KNeighborsClassifier import numpy as np from sklearn.model_selection import train_test_split from matplotlib.colors import ListedColormap import pandas as pd from sklearn.metrics import accuracy_score import seaborn as sns import matplotlib.pyplot as plt import matplotlib.lines as mlines import plotly_express as px

data = pd.read_csv("data.csv") data = data[['Gender', 'Age', 'Employment status','Attitudes towards the LIVE that sells products ']] data.shape

feature = data[['Gender', 'Age', 'Employment status']].values label = data['Attitudes towards the LIVE that sells products '].values

#knn算法预测 X_train, X_test, y_train, y_test = train_test_split(feature,label,test_size=0.9) knn = KNeighborsClassifier() knn.fit(X_train, y_train) predict_y = knn.predict(X_test) print("KNN 准确率", accuracy_score(y_test, predict_y))

#选取9个近邻 X_train_new = X_train[:18] X_train_validation = X_train[18:] for k in range(1, 15, 2): knn = KNeighborsClassifier(n_neighbors=k) knn.fit(X_train, y_train) predict_y = knn.predict(X_test) print("K为%s的准确率" % k, accuracy_score(y_test, predict_y))

returnMat, classLabelVector = np.array(data[['Gender', 'Age', 'Employment status']]),data['Attitudes towards the LIVE that sells products ']

def showdatas(returnMat, classLabelVector): plt.rcParams['axes.unicode_minus']=False # 用来正常显示负号 # 设置2*2的画布 fig,axs = plt.subplots(nrows=2, ncols=2,sharex=False,sharey=False,figsize=(13,8)) # 特征数据长度 numberOfLabels = len(classLabelVector) # 设置空标签来存储 LabelsColors = [] for i in classLabelVector: # 对每个标签遍历，并且赋上对应的颜色 if i == 1: LabelsColors.append('blue') if i == 2: LabelsColors.append('yellow') if i == 3: LabelsColors.append('red') # 画出散点图：散点大小15 透明度0.5 axs[0][0].scatter(x=returnMat[:,0],y=returnMat[:,1],color=LabelsColors,s=15,alpha=0.5) # gender, age axs0_title_text = axs[0][0].set_title(u'gender,age') axs0_xlabel_text = axs[0][0].set_xlabel(u'gender') axs0_ylabel_text = axs[0][0].set_ylabel(u'age') plt.setp(axs0_title_text, size=9, weight='bold', color='red') plt.setp(axs0_xlabel_text, size=7, weight='bold', color='black') plt.setp(axs0_ylabel_text, size=7, weight='bold', color='black') # gender, employment axs[0][1].scatter(x=returnMat[:,0],y=returnMat[:,2],color=LabelsColors,s=15,alpha=0.5) # 设置标题、x轴、y轴 axs1_title_text = axs[0][1].set_title(u'gender,employment') axs1_xlabel_text = axs[0][1].set_xlabel(u'gender') axs1_ylabel_text = axs[0][1].set_ylabel(u'employment') plt.setp(axs1_title_text, size=9, weight='bold', color='red') plt.setp(axs1_xlabel_text, size=7, weight='bold', color='black') plt.setp(axs1_ylabel_text, size=7, weight='bold', color='black') # age, employment axs[1][0].scatter(x=returnMat[:,1],y=returnMat[:,2],color=LabelsColors,s=15,alpha=0.5) # 设置标题、x轴、y轴 axs2_title_text = axs[1][0].set_title(u'age,employment') axs2_xlabel_text = axs[1][0].set_xlabel(u'age') axs2_ylabel_text = axs[1][0].set_ylabel(u'employment') plt.setp(axs2_title_text, size=9, weight='bold', color='red') plt.setp(axs2_xlabel_text, size=7, weight='bold', color='black') plt.setp(axs2_ylabel_text, size=7, weight='bold', color='black') # 设置图例 like = mlines.Line2D([], [], color='blue', marker='.', markersize=6, label='like') indifferent = mlines.Line2D([], [], color='yellow', marker='.', markersize=6, label='indifferent') dislike = mlines.Line2D([], [], color='red', marker='.', markersize=6, label='dont like') #添加图例 axs[0][0].legend(handles=[like,indifferent,dislike]) axs[0][1].legend(handles=[like,indifferent,dislik]) axs[1][0].legend(handles=[like,indifferent,dislik]) #显示图片 plt.show() showdatas(returnMat, classLabelVector)

#特征归一化 def minmax(dataSet): minD = dataSet.min() maxD = dataSet.max() normSet = (dataSet - minD) / (maxD - minD) return normSet dataNew = pd.concat([minmax(data.iloc[:,:3]), data.iloc[:,3]], axis=1) dataNew

returnMat1, classLabelVector1 = np.array(dataNew[['Gender', 'Age', 'Employment status']]),dataNew['Attitudes towards the LIVE that sells products '] classLabelVector1 showdatas(returnMat1, classLabelVector1)

feature1 = dataNew[['Gender', 'Age', 'Employment status']].values label1 = dataNew['Attitudes towards the LIVE that sells products '].values

def randSplit(dataSet, rate=0.9): n = dataSet.shape[0] # [1000,4] 取出1000，实际上就是data的长度len(data) m = int(n * rate) # 取出前90% train = dataSet.iloc[:m,:] # 前90%的行，所有列 test = dataSet.iloc[m:,:] # 后面的行及所有列 test.index = range(test.shape[0]) # 测试集test的上索引需要重置 return train,test # 调用函数 train,test = randSplit(dataNew) def dataClassify(train,test,k): n = train.shape[1] - 1 # train除去标签的所有列 m = test.shape[0] # test的行数 result = [] # 存放最终的结果 for i in range(m): dist = list((((train.iloc[:,:n] - test.iloc[i,:n]) ** 2).sum(1))**0.5) # 计算训练集中的每个数据和测试集中某个数据的欧氏距离 dist_l = pd.DataFrame({'dist':dist, 'labels':(train.iloc[:,n])}) # 计算出来的距离和对应训练集的标签构成DF型数据 dr = dist_l.sort_values(by='dist')[:k] # 根据标签排序取出前k个数据 re = dr.loc[:,'labels'].value_counts() # 前k个数据中统计每个标签出现的票数，票数高的则为测试集数据的标签 result.append(re.index[0]) # re.index[0]表示票数最高的分类 result = pd.Series(result) test['predict'] = result # 测试集中添加预测的结果 acc = (test.iloc[:,-1] == test.iloc[:,-2]).mean() # 原始结果和预测结果的对比 print(f"模型预测准确率{acc}") return test dataClassify(train,test,9)