# -*- coding:utf-8 -*-
import pandas as pd
import numpy as np
import sys
from sklearn.model_selection import train_test_split
from sklearn.neighbors import KNeighborsClassifier, KNeighborsRegressor
from sklearn.metrics import accuracy_score, recall_score, precision_score, r2_score
from sklearn.preprocessing import LabelEncoder
# 1. 加载数据
df = pd.read_csv('iris.data', header=None, sep=",")
# 2. 数据预览
print(df.head())
print(df.shape)
# 3. 数据预处理
X = df.iloc[:, 0:4]
Y = df.iloc[:, -1]
# 4. 数据清洗
label_encoder = LabelEncoder()
Y = label_encoder.fit_transform(Y)
# 5. 数据分割
x_train, x_test, y_train, y_test = train_test_split(X, Y, train_size=0.8, random_state=1)
print("训练数据X的shape:{}, 以及类型:{}".format(x_train.shape, type(x_train)))
print("测试数据X的shape:{}, 以及类型:{}".format(x_test.shape, type(x_test)))
print("训练数据Y的shape:{}，以及类型:{}".format(y_train.shape, type(y_train)))
# 6. 模型对象的构建
KNN = KNeighborsClassifier(n_neighbors=5, weights='uniform', algorithm='kd_tree')
# 7. 模型的训练
KNN.fit(x_train, y_train)
# 8. 模型效果评估
train_predict = KNN.predict(x_train)
test_predict = KNN.predict(x_test)
print("KNN算法：测试集上的效果(F1-score):{}".format(KNN.score(x_test, y_test)))
print("KNN算法：训练集上的效果(F1-score):{}".format(KNN.score(x_train, y_train)))
print("KNN算法：训练集上的效果(准确率):{}".format(accuracy_score(y_true=y_train, y_pred=train_predict)))
# 9.模型的保存与加载(pip install joblib)
import joblib
# 保存模型
joblib.dump(KNN, "./knn.m")
# 加载模型
# joblib.load("./knn.m")
