diff --git a/function.py b/function.py index 40d4ec6..9b40e50 100644 --- a/function.py +++ b/function.py @@ -27,85 +27,129 @@ import os """ id = "202210423122" name = "焦国波" - - - -参数说明: -- input_vector (numpy array): 输入待分类的样本。 -- dataSet (numpy array): 训练数据集。 -- labels (list or numpy array): 训练集对应的标签。 -- k (int): kNN 算法参数,请**自行设置默认值**,评测时仅会使用其默认值,不修改默认值将**无法通过**评测,**成绩为零**。 -返回值: -- int: 预测的分类标签。 -""" -def load_data(directory): - """ - 加载数据集,将图像数据转换为向量形式,并提取标签。 - - 参数: - - directory (str): 数据集目录路径。 - - 返回值: - - images (numpy array): 图像数据。 - - labels (list): 标签列表。 - """ - images = [] - labels = [] - for filename in os.listdir(directory): - if filename.endswith(".png"): - label = int(filename.split('_')[0]) - image = np.array([int(pixel) for pixel in open(os.path.join(directory, filename), 'rb').read()]) - images.append(image) - labels.append(label) - return np.array(images), np.array(labels) - def classify(input_vector, dataSet, labels, k=3): """ - kNN 分类算法实现。 - - 参数: + KNN 分类算法。 + + Args: - input_vector (numpy array): 输入待分类的样本。 - dataSet (numpy array): 训练数据集。 - labels (list or numpy array): 训练集对应的标签。 - - k (int): kNN 算法参数,默认值为 3。 - - 返回值: + - k (int): kNN 算法参数,默认设置为3。 + + Returns: - int: 预测的分类标签。 """ - # 计算输入向量与数据集中每个样本的欧式距离 - distances = np.sqrt(np.sum((dataSet - input_vector) ** 2, axis=1)) + # 计算输入向量与训练集样本的欧氏距离 + distances = np.linalg.norm(dataSet - input_vector, axis=1) + + # 按距离排序并获取前k个距离最小的索引 + sorted_indices = distances.argsort()[:k] + + # 统计前k个邻居的标签 + nearest_labels = [labels[i] for i in sorted_indices] + + # 返回出现次数最多的标签 + predicted_label = max(set(nearest_labels), key=nearest_labels.count) + + return predicted_label + + +def load_dataset(directory): + """ + 从文件夹中加载数据,若不自行正确实现,则无法通过评测,**不记录成绩**。 + + Args: + - directory (str): 数据文件夹路径。 + + Returns: + - np.array: 特征数据矩阵。 + - np.array: 标签数组。 + """ + file_list = os.listdir(directory) + num_files = len(file_list) - # 获取 k 个最近邻的索引 - sortedDistIndices = distances.argsort() + # 初始化数据矩阵和标签数组 + data_matrix = np.zeros((num_files, 1024)) + labels = [] - # 获取 k 个最近邻的标签 - classCount = {} - for i in range(k): - voteLabel = labels[sortedDistIndices[i]] - classCount[voteLabel] = classCount.get(voteLabel, 0) + 1 + for i, file_name in enumerate(file_list): + # 文件名格式:label_index.txt,例如 0_1.txt,标签为 0 + label = int(file_name.split('_')[0]) + labels.append(label) + data_matrix[i, :] = img_to_vector(os.path.join(directory, file_name)) - # 多数表决,返回出现次数最多的标签 - sortedClassCount = sorted(classCount.items(), key=lambda x: x[1], reverse=True) - return sortedClassCount[0][0] + return data_matrix, np.array(labels) + + +def img_to_vector(filename): + """ + 将二值图像转换为向量。 + 该部分应为 load_dataset 函数的辅助子函数,可单独在这里实现,也可直接集成至 load_dataset 函数中。 + + Args: + - filename (str): 图像文件路径。 + + Returns: + - np.array: 二值图像向量 (1x1024)。 + """ + return_vector = np.zeros((1, 1024)) # 创建1x1024的零向量 + with open(filename) as file: + for i in range(32): # 图像为32x32 + line_str = file.readline() + for j in range(32): + return_vector[0, 32 * i + j] = int(line_str[j]) # 转换为1x1024向量 + return return_vector + -# 以下代码用于测试和演示,实际使用时请删除或注释掉 if __name__ == "__main__": - # 加载训练集和测试集 - X_train, y_train = load_data('trainingDigits') - X_test, y_test = load_data('testDigits') - - # 测试分类函数 - test_samples = X_test[:5] - for sample in test_samples: - predicted_label = classify(sample, X_train, y_train) - print(f"Input Vector: {sample}, Predicted Label: {predicted_label}") + # 测试代码示例,加载训练集并进行分类 + train_data, train_labels = load_dataset("trainingDigits") + test_data = ( + "00000000000000110000000000000000" + "00000000000011111000000000000000" + "00000000000011111100000000000000" + "00000000001111111110000000000000" + "00000000001111111110000000000000" + "00000000001111111100000000000000" + "00000000011111111000000000000000" + "00000000111111111000000000000000" + "00000000111111110000000000000000" + "00000000111111110000000000000000" + "00000000111111110000000000000000" + "00000000111111110000000000000000" + "00000000111111100000000000000000" + "00000001111111100000000000000000" + "00000001111111100000000000000000" + "00000001111111100000000000000000" + "00000001111111111111000000000000" + "00000001111111111111111000000000" + "00000001111111111111111100000000" + "00000011111111111111111110000000" + "00000011111111111111111110000000" + "00000001111111111111111110000000" + "00000001111111111111111110000000" + "00000000111111111100111111100000" + "00000000111111111000111111100000" + "00000000011111111001111111100000" + "00000000011111111111111111000000" + "00000000001111111111111111000000" + "00000000000111111111111111000000" + "00000000000011111111111100000000" + "00000000000001111111110000000000" "00000000000000000111100000000000" ) - predicted_label = classify( - np.array([int(char) for char in test_data]).reshape(1, 1024), - train_data, - train_labels, - ) + + # 将测试数据转换为向量 + test_vector = np.array([int(char) for char in test_data]).reshape(1, 1024) + + # 使用 kNN 进行分类 + predicted_label = classify(test_vector, train_data, train_labels) print(f"预测的标签: {predicted_label}") + + + + + \ No newline at end of file