生成一个根据提供的手机相册分类数据集，制作成训练图像数据集、测试图像数据集、训练标签数据集和测试标签数据集。将四部分分别保存成numpy文件的代码

以下是一个将提供的手机相册分类数据集制作成训练图像数据集、测试图像数据集、训练标签数据集和测试标签数据集，并将四部分分别保存成numpy文件的代码示例：

import os
import numpy as np
from PIL import Image

# 设置数据集路径
data_path = '/path/to/dataset'

# 设置训练集、测试集的比例
train_ratio = 0.8
test_ratio = 0.2

# 获取所有图像文件的路径和标签
image_paths = []
labels = []

for label_name in os.listdir(data_path):
    label_path = os.path.join(data_path, label_name)
    if os.path.isdir(label_path):
        for image_name in os.listdir(label_path):
            image_path = os.path.join(label_path, image_name)
            image_paths.append(image_path)
            labels.append(label_name)

# 将数据集随机打乱
indices = np.arange(len(image_paths))
np.random.shuffle(indices)
image_paths = np.array(image_paths)[indices]
labels = np.array(labels)[indices]

# 划分数据集
train_size = int(len(image_paths) * train_ratio)
test_size = int(len(image_paths) * test_ratio)
train_image_paths = image_paths[:train_size]
train_labels = labels[:train_size]
test_image_paths = image_paths[train_size:train_size+test_size]
test_labels = labels[train_size:train_size+test_size]

# 加载图像并将其保存为numpy数组
def load_and_save_images(image_paths, save_path):
    images = []
    for image_path in image_paths:
        image = Image.open(image_path)
        image = image.resize((224, 224))  # 调整图像大小
        image = np.array(image)
        images.append(image)
    images = np.array(images)
    np.save(save_path, images)

# 加载标签并将其保存为numpy数组
def load_and_save_labels(labels, save_path):
    label_to_index = {label: index for index, label in enumerate(np.unique(labels))}
    indices = np.array([label_to_index[label] for label in labels])
    np.save(save_path, indices)

# 保存训练集和测试集的图像和标签
train_image_save_path = '/path/to/save/train_images.npy'
test_image_save_path = '/path/to/save/test_images.npy'
train_label_save_path = '/path/to/save/train_labels.npy'
test_label_save_path = '/path/to/save/test_labels.npy'

load_and_save_images(train_image_paths, train_image_save_path)
load_and_save_images(test_image_paths, test_image_save_path)
load_and_save_labels(train_labels, train_label_save_path)
load_and_save_labels(test_labels, test_label_save_path)

上述代码将数据集随机打乱后，按照指定比例划分为训练集和测试集，并将图像和标签分别保存为numpy数组文件。其中，`load_and_save_images`函数用于加载图像并将其保存为numpy数组，`load_and_save_labels`函数用于加载标签并将其保存为numpy数组。在使用时需要将代码中的`/path/to/dataset`、`/path/to/save/train_images.npy`、`/path/to/save/test_images.npy`、`/path/to/save/train_labels.npy`和`/path/to/save/test_labels.npy`替换为实际的路径。