1.获取数据
import pandas as pd
import numpy as np
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import StandardScaler
from sklearn.linear_model import LogisticRegression
data = pd.read_csv(r'data\distance.csv')
data.head()
2. 基本数据处理
# 2.1 缺失值处理
data = data.replace(to_replace="?", value=np.NaN) # 把data中的 to_replece 值 替换为 value.
data = data.dropna() # 删除有空值的行,默认 axis=0
data.head()
3.确定特征值,目标值
data.columns # 所有的列名
Index(['Unnamed: 0', 'A0', 'A1', 'A2', 'A3', 'x', 'y', 'z', 'label'], dtype='object')
x = data.iloc[:, 1:8]
x.head()
y = data["label"]
y.head()
0 1
1 1
2 1
3 1
4 1
Name: label, dtype: int64
x.head()
x.shape[0]*0.75
486.0
4.分割数据
X_train, X_test, y_train, y_test = train_test_split(x, y, random_state=22) # 训练集和测试集按照 0.75 : 0.25 (随机)
X_train
X_test
y_train
327 0
56 1
242 1
5 1
449 0
..
491 0
502 0
358 0
356 0
132 1
Name: label, Length: 486, dtype: int64
y_test
623 0
389 0
551 0
617 0
130 1
..
568 0
427 0
485 0
116 1
148 1
Name: label, Length: 162, dtype: int64
5.特征工程(标准化)
transfer = StandardScaler() # 实例化对象
X_train = transfer.fit_transform(X_train) # 标准化
X_test = transfer.fit_transform(X_test)
X_train
array([[-1.43796501, 0.86810151, -0.54520456, ..., -1.55143148,
-0.39370337, -1.39230246],
[ 0.05371734, -1.15687309, 1.03849334, ..., 0.73802915,
-0.77946917, -1.39230246],
[ 2.07895166, 0.61764551, 0.48481435, ..., 1.50118269,
1.53512564, 0.55690142],
...,
[-1.28928037, -0.47243389, 0.55503119, ..., -0.40670117,
-1.16523498, -1.39230246],
[-0.83365733, -0.13412409, -0.09032518, ..., -0.40670117,
-0.39370337, -1.39230246],
[-0.4273051 , -1.03258542, 0.9046769 , ..., 0.35645237,
-0.77946917, -0.39392974]])
X_test
array([[ 0.20623287, -1.62114508, 1.00301367, ..., 0.89636437,
-0.76011455, 1.22013336],
[ 1.18389765, -0.15251848, 0.47513551, ..., 1.30426052,
0.80852949, -1.42227529],
[ 1.27129309, -0.2730585 , 0.35857018, ..., 1.30426052,
0.80852949, 0.51234533],
...,
[ 1.92764066, 0.57275967, 0.58383134, ..., 1.71215667,
1.59285151, -0.43137205],
[-1.29682755, -0.64814879, 0.99497082, ..., -0.32732407,
-1.54443657, -0.43137205],
[ 0.74174642, -1.10997832, 0.76310322, ..., 1.30426052,
0.02420747, -0.43137205]])
6.机器学习(逻辑回归)
estimator = LogisticRegression()
estimator.fit(X_train, y_train)
estimator
LogisticRegression()
7.模型评估
y_predict = estimator.predict(X_test)
y_predict
array([0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 1, 1, 0, 1, 1, 1, 0, 0,
1, 1, 0, 1, 1, 1, 1, 0, 1, 0, 0, 1, 1, 1, 0, 1, 1, 0, 1, 0, 0, 1,
0, 1, 1, 1, 0, 0, 0, 0, 1, 0, 1, 0, 1, 0, 1, 1, 0, 0, 0, 0, 0, 1,
0, 1, 0, 1, 0, 1, 1, 0, 0, 1, 1, 0, 1, 0, 1, 1, 1, 0, 0, 1, 0, 1,
0, 1, 0, 1, 0, 0, 0, 1, 1, 0, 1, 1, 0, 0, 1, 0, 1, 0, 1, 0, 1, 0,
1, 1, 1, 1, 0, 1, 0, 0, 1, 1, 1, 0, 0, 1, 1, 1, 1, 1, 1, 1, 0, 0,
0, 1, 1, 1, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 1, 1, 0, 0, 0,
1, 1, 1, 1, 0, 1, 1, 0], dtype=int64)
estimator.score(X_test, y_test)
0.5864197530864198
X_test
array([[ 0.20623287, -1.62114508, 1.00301367, ..., 0.89636437,
-0.76011455, 1.22013336],
[ 1.18389765, -0.15251848, 0.47513551, ..., 1.30426052,
0.80852949, -1.42227529],
[ 1.27129309, -0.2730585 , 0.35857018, ..., 1.30426052,
0.80852949, 0.51234533],
...,
[ 1.92764066, 0.57275967, 0.58383134, ..., 1.71215667,
1.59285151, -0.43137205],
[-1.29682755, -0.64814879, 0.99497082, ..., -0.32732407,
-1.54443657, -0.43137205],
[ 0.74174642, -1.10997832, 0.76310322, ..., 1.30426052,
0.02420747, -0.43137205]])
y_test
623 0
389 0
551 0
617 0
130 1
..
568 0
427 0
485 0
116 1
148 1
Name: label, Length: 162, dtype: int64
from sklearn.metrics import precision_score, recall_score, f1_score
precision = precision_score(np.array(y_test), np.array(y_predict))
recall = recall_score(y_test, y_predict)
f1 = f1_score(y_test, y_predict)
print(precision)
print(recall)
print(f1)
0.6049382716049383
0.5833333333333334
0.5939393939393939
总结
逻辑回归的准确率、召回率和F1-socre的分数相对较大,模型预测的结果不好。