metodi/app/labs/lab7.py

import numpy as np
import matplotlib.pyplot as plt
import sklearn.metrics
from sklearn.datasets import load_iris
from sklearn.model_selection import train_test_split
from sklearn.neighbors import KNeighborsClassifier
import matplotlib

class lab7:

    def __init__(self):
        pass

    def main(self):
        matplotlib.use('TkAgg')

        # Загружаем данные ирисов
        data = load_iris()
        X = data.data  # Признаки ирисов
        Y = data.target  # Сорта ирисов
        label = Y  # Используем Y в качестве меток классов для классификации
        Y_str = data.target_names  # Названия сортов ирисов

        # setosa
        setosa_inds = Y == np.where(Y_str == "setosa")[0][0]
        setosa_data = X[setosa_inds, :]
        # versicolor
        versicolor_inds = Y == np.where(Y_str == "versicolor")[0][0]
        versicolor_data = X[versicolor_inds, :]
        # virginica
        virginica_inds = Y == np.where(Y_str == "virginica")[0][0]
        virginica_data = X[virginica_inds, :]

        # Определяем функцию классификации
        def classification(data, label, k):
            X_train, X_test, y_train, y_test = train_test_split(data, label, random_state=10)  #Разделение данных
            knn = KNeighborsClassifier(n_neighbors=k, metric='euclidean')  # Создание KNN-классификатора
            knn.fit(X_train, y_train)
            predicted_data = knn.predict(X_test)
            report = sklearn.metrics.classification_report(y_test, predicted_data)

        # Пример вызова функции классификации
        setosa_data = sklearn.preprocessing.normalize(setosa_data)
        versicolor_data = sklearn.preprocessing.normalize(versicolor_data)
        virginica_data = sklearn.preprocessing.normalize(virginica_data)
        data_all = np.vstack((setosa_data,versicolor_data,virginica_data))

        fig, axs = plt.subplots(2, 3, figsize=(12,8),dpi=160)
        axs[0, 0].scatter(data_all[:50, 0], data_all[:50, 1], marker='*')
        axs[0, 0].scatter(data_all[50:100, 0], data_all[50:100, 1], marker='^')
        axs[0, 0].scatter(data_all[100:, 0], data_all[100:, 1], marker='+')
        axs[0, 0].title.set_text('features 1 and 2')

        axs[0, 1].scatter(data_all[:50, 0], data_all[:50, 2], marker='*')
        axs[0, 1].scatter(data_all[50:100, 0], data_all[50:100, 2], marker='^')
        axs[0, 1].scatter(data_all[100:, 0], data_all[100:, 2], marker='+')
        axs[0, 1].title.set_text('features 1 and 3')

        axs[0, 2].scatter(data_all[:50, 0], data_all[:50, 3], marker='*')
        axs[0, 2].scatter(data_all[50:100, 0], data_all[50:100, 3], marker='^')
        axs[0, 2].scatter(data_all[100:, 0], data_all[100:, 3], marker='+')
        axs[0, 2].title.set_text('features 1 and 4')

        axs[1, 0].scatter(data_all[:50, 1], data_all[:50, 2], marker='*',)
        axs[1, 0].scatter(data_all[50:100, 1], data_all[50:100, 2], marker='^')
        axs[1, 0].scatter(data_all[100:, 1], data_all[100:, 2], marker='+')
        axs[1, 0].title.set_text('features 2 and 3')

        axs[1, 1].scatter(data_all[:50, 1], data_all[:50, 3], marker='*')
        axs[1, 1].scatter(data_all[50:100, 1], data_all[50:100, 3], marker='^')
        axs[1, 1].scatter(data_all[100:, 1], data_all[100:, 3], marker='+')
        axs[1, 1].title.set_text('features 2 and 4')

        axs[1, 2].scatter(data_all[:50, 2], data_all[:50, 3], marker='*')
        axs[1, 2].scatter(data_all[50:100, 2], data_all[50:100, 3], marker='^')
        axs[1, 2].scatter(data_all[100:, 2], data_all[100:, 3], marker='+')
        axs[1, 2].title.set_text('features 3 and 4')
        plt.show()

        data = np.hstack((data_all[:,0].reshape(150,1),data_all[:,1].reshape(150,1)))
        print('Количество соседей = 2:')
        classification(data,label,2)
        print('Количество соседей = 5:')
        classification(data,label,5)
        print('Количество соседей = 20:')
        classification(data,label,20)
        print('Количество соседей = 70:')
        classification(data,label,70)
        print('///////////////////////////////////////////////////////////////////')
        data = np.hstack((data_all[:,0].reshape(150,1),data_all[:,2].reshape(150,1)))
        print('Количество соседей = 2:')
        classification(data,label,2)
        print('Количество соседей = 5:')
        classification(data,label,5)
        print('Количество соседей = 20:')
        classification(data,label,20)
        print('Количество соседей = 70:')
        classification(data,label,70)
        print('///////////////////////////////////////////////////////////////////')
        data = np.hstack((data_all[:,0].reshape(150,1),data_all[:,3].reshape(150,1)))
        print('Количество соседей = 2:')
        classification(data,label,2)
        print('Количество соседей = 5:')
        classification(data,label,5)
        print('Количество соседей = 20:')
        classification(data,label,20)
        print('Количество соседей = 70:')
        classification(data,label,70)
        print('///////////////////////////////////////////////////////////////////')
        data = np.hstack((data_all[:,1].reshape(150,1),data_all[:,2].reshape(150,1)))
        print('Количество соседей = 2:')
        classification(data,label,2)
        print('Количество соседей = 5:')
        classification(data,label,5)
        print('Количество соседей = 20:')
        classification(data,label,20)
        print('Количество соседей = 70:')
        classification(data,label,70)
        print('///////////////////////////////////////////////////////////////////')
        data = np.hstack((data_all[:,1].reshape(150,1),data_all[:,3].reshape(150,1)))
        print('Количество соседей = 2:')
        classification(data,label,2)
        print('Количество соседей = 5:')
        classification(data,label,5)
        print('Количество соседей = 20:')
        classification(data,label,20)
        print('Количество соседей = 70:')
        classification(data,label,70)
        print('///////////////////////////////////////////////////////////////////')
        data = np.hstack((data_all[:,2].reshape(150,1),data_all[:,3].reshape(150,1)))
        print('Количество соседей = 2:')
        classification(data,label,2)
        print('Количество соседей = 5:')
        classification(data,label,5)
        print('Количество соседей = 20:')
        classification(data,label,20)
        print('Количество соседей = 70:')
        classification(data,label,70)


if __name__ == "__main__":
    l7 = lab7()
    l7.main()