Python 74_ Logistic Regression 로지스틱 회귀

from sklearn import datasets

import pandas as pd

import seaborn as sns

import matplotlib.pyplot as plt

iris = datasets.load_iris()

X = iris['data'] # iris.data

y = iris['target'] # iris.target

features = iris['feature_names'] # iris.feature_names

print(features)

['sepal length (cm)', 'sepal width (cm)', 'petal length (cm)', 'petal width (cm)']

iris_df = pd.DataFrame(X,

columns=['sepal_length', 'sepal_width',

'petal_length', 'petal_width'])

# 데이터 프레임에 컬럼(변수, 특성)을 추가

iris_df['species'] = y

print(iris_df.iloc[:5, :])

print(iris_df.describe())

   sepal_length  sepal_width  petal_length  petal_width  species

0           5.1          3.5           1.4          0.2        0

1           4.9          3.0           1.4          0.2        0

2           4.7          3.2           1.3          0.2        0

3           4.6          3.1           1.5          0.2        0

4           5.0          3.6           1.4          0.2        0

       sepal_length  sepal_width  petal_length  petal_width     species

count    150.000000   150.000000    150.000000   150.000000  150.000000

mean       5.843333     3.057333      3.758000     1.199333    1.000000

std        0.828066     0.435866      1.765298     0.762238    0.819232

min        4.300000     2.000000      1.000000     0.100000    0.000000

25%        5.100000     2.800000      1.600000     0.300000    0.000000

50%        5.800000     3.000000      4.350000     1.300000    1.000000

75%        6.400000     3.300000      5.100000     1.800000    2.000000

max        7.900000     4.400000      6.900000     2.500000    2.000000

데이터의 개괄적인 특징 파악하기

sns.pairplot(iris_df, hue='species', vars=['sepal_length', 'sepal_width', 'petal_length', 'petal_width'])

plt.show()

# 데이터(X)와 타겟(y)을 학습/검증 세트로 분리

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=1217)

# 분류 알고리즘 중에서 Logistic Regression을 선택

log_reg = LogisticRegression()

# 모델 적합(fitting)/학습(training)

log_reg.fit(X_train, y_train)

# 예측/테스트

predictions = log_reg.predict(X_test)

print('y true:', y_test)

print('y pred:', predictions)

y true: [1 2 2 0 1 2 1 2 0 0 1 1 2 1 1 0 2 0 1 0 2 0 0 1 1 1 1 0 2 2]

y pred: [1 2 2 0 1 2 1 2 0 0 1 1 2 1 1 0 2 0 1 0 2 0 0 2 1 1 1 0 2 1]

# 성능 측정

print(confusion_matrix(y_test, predictions))

print(classification_report(y_test,predictions))

[[ 9  0  0]

 [ 0 11  1]

 [ 0  1  8]]

              precision    recall  f1-score   support

           0       1.00      1.00      1.00         9

           1       0.92      0.92      0.92        12

           2       0.89      0.89      0.89         9

    accuracy                           0.93        30

   macro avg       0.94      0.94      0.94        30

weighted avg       0.93      0.93      0.93        30