【knn临近算法】预测年收入

import numpy as np
import pandas as pd 
from sklearn.neighbors import KNeighborsClassifier # 导入knn学习包

salary = pd.read_csv("./adults.txt")
salary.head()

#样本数据的提取
y = salary['salary']
X = salary.iloc[:,[0,1,3,5,6,8,9,-2,-3]]
X.head()

# 数据去重 用于下面函数的理解
u = X['occupation'].unique()
u

# np.argwhere此方法找到对应名称的索引
np.argwhere(u == "Exec-managerial")[0,0]  # np.argwhere

X.columns[2:-1]

# 2:-1列的所有数据映射
for col in X.columns[1:-1]:  # 遍历所有类名
    
    u = X[col].unique()  #  类似上面的u = X['occupation'].unique()  得出每个分类下面的种类名称
#     print(col)

    def convert(x):  # 将上面得出的u 进行索引映射
#         print(x)
        return np.argwhere(u == x)[0,0]  #  将上面得出的u 进行索引映射
    
    X[col] = X[col].map(convert)  # 将上面得出的u 进行索引映射
    
X.head()

# 切分训练集跟测试集
from sklearn.model_selection import train_test_split

X_train,X_test,y_train,y_test = train_test_split(X,y,test_size = 0.2)  # 切分
print(X_train.shape,X_test.shape,y_train.shape,y_test.shape)

# 预测测试
knn = KNeighborsClassifier(n_neighbors=8)

knn.fit(X_train,y_train)  # 计算公式

y_ = knn.predict(X_test) # 预测值

from sklearn.metrics import accuracy_score # 计算分类预测的准确率

# 求出预测准确率
accuracy = accuracy_score(y_test, y_)

print("预测准确率: ", accuracy)

总结：难度在于数据的预处理