From 048088cd73bc45ab6f35e41b6a89444a20f81a6b Mon Sep 17 00:00:00 2001
From: Fisherlen <fisherlen.yu@163.com>
Date: Sun, 15 Mar 2026 12:29:29 +0800
Subject: [PATCH] Add dataset download script for Hugging Face

---
 download_datasets.py | 118 +++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 118 insertions(+)
 create mode 100644 download_datasets.py

diff --git a/download_datasets.py b/download_datasets.py
new file mode 100644
index 000000000..101a040f6
--- /dev/null
+++ b/download_datasets.py
@@ -0,0 +1,118 @@
+"""
+下载 Time-Series-Library 数据集从 Hugging Face
+"""
+import os
+from datasets import load_dataset
+import pandas as pd
+
+HUGGINGFACE_REPO = "thuml/Time-Series-Library"
+DATASET_ROOT = "./Time-Series-Library/dataset"
+
+# 根据 Hugging Face 实际可用的配置
+AVAILABLE_CONFIGS = [
+    # ETT 数据集
+    "ETTh1", "ETTh2", "ETTm1", "ETTm2",
+    # 主要预测数据集
+    "electricity", "traffic", "weather", "exchange_rate", "national_illness",
+    # M4 数据集
+    "m4-yearly", "m4-quarterly", "m4-monthly", "m4-weekly", "m4-daily", "m4-hourly",
+    # 分类数据集 (UEA)
+    "EthanolConcentration", "FaceDetection", "Handwriting", "Heartbeat",
+    "JapaneseVowels", "PEMS-SF", "SelfRegulationSCP1", "SelfRegulationSCP2",
+    "SpokenArabicDigits", "UWaveGestureLibrary",
+    # 异常检测数据集
+    "SMD-data", "SMD-label", "MSL-data", "MSL-label",
+    "SMAP-data", "SMAP-label", "PSM-data", "PSM-label", "SWaT"
+]
+
+def download_dataset(config_name, output_dir):
+    """下载单个数据集并保存为CSV"""
+    try:
+        print(f"正在下载数据集: {config_name}")
+        ds = load_dataset(HUGGINGFACE_REPO, name=config_name)
+
+        # 保存训练集为CSV
+        df = ds["train"].to_pandas()
+        output_path = os.path.join(output_dir, f"{config_name}.csv")
+        df.to_csv(output_path, index=False)
+        print(f"  ✓ 已保存到: {output_path}")
+        return True
+    except Exception as e:
+        print(f"  ✗ 下载失败 {config_name}: {e}")
+        return False
+
+def create_directory_structure():
+    """创建数据集目录结构"""
+    dirs = [
+        os.path.join(DATASET_ROOT, "ETT-small"),
+        os.path.join(DATASET_ROOT, "electricity"),
+        os.path.join(DATASET_ROOT, "exchange_rate"),
+        os.path.join(DATASET_ROOT, "traffic"),
+        os.path.join(DATASET_ROOT, "weather"),
+        os.path.join(DATASET_ROOT, "illness"),
+        os.path.join(DATASET_ROOT, "m4"),
+        os.path.join(DATASET_ROOT, "anomaly"),
+        os.path.join(DATASET_ROOT, "classification"),
+    ]
+    for d in dirs:
+        os.makedirs(d, exist_ok=True)
+        print(f"创建目录: {d}")
+
+def main():
+    print("=" * 60)
+    print("Time-Series-Library 数据集下载工具")
+    print("=" * 60)
+
+    # 创建目录结构
+    create_directory_structure()
+
+    # 下载 ETT 数据集
+    print("\n[1/5] 下载 ETT 数据集...")
+    ett_configs = ["ETTh1", "ETTh2", "ETTm1", "ETTm2"]
+    for config in ett_configs:
+        download_dataset(config, os.path.join(DATASET_ROOT, "ETT-small"))
+
+    # 下载主要预测数据集
+    print("\n[2/5] 下载其他预测数据集...")
+    forecast_configs = {
+        "electricity": "electricity",
+        "exchange_rate": "exchange_rate",
+        "traffic": "traffic",
+        "weather": "weather",
+        "national_illness": "illness",
+    }
+    for config, folder in forecast_configs.items():
+        download_dataset(config, os.path.join(DATASET_ROOT, folder))
+
+    # 下载 M4 数据集
+    print("\n[3/5] 下载 M4 数据集...")
+    m4_configs = ["m4-yearly", "m4-quarterly", "m4-monthly", "m4-weekly", "m4-daily", "m4-hourly"]
+    for config in m4_configs:
+        download_dataset(config, os.path.join(DATASET_ROOT, "m4"))
+
+    # 下载分类数据集
+    print("\n[4/5] 下载分类数据集...")
+    classification_configs = [
+        "EthanolConcentration", "FaceDetection", "Handwriting", "Heartbeat",
+        "JapaneseVowels", "PEMS-SF", "SelfRegulationSCP1", "SelfRegulationSCP2",
+        "SpokenArabicDigits", "UWaveGestureLibrary"
+    ]
+    for config in classification_configs:
+        download_dataset(config, os.path.join(DATASET_ROOT, "classification"))
+
+    # 下载异常检测数据集
+    print("\n[5/5] 下载异常检测数据集...")
+    anomaly_configs = [
+        "SMD-data", "SMD-label", "MSL-data", "MSL-label",
+        "SMAP-data", "SMAP-label", "PSM-data", "PSM-label", "SWaT"
+    ]
+    for config in anomaly_configs:
+        download_dataset(config, os.path.join(DATASET_ROOT, "anomaly"))
+
+    print("\n" + "=" * 60)
+    print("数据集下载完成!")
+    print(f"数据集保存在: {os.path.abspath(DATASET_ROOT)}")
+    print("=" * 60)
+
+if __name__ == "__main__":
+    main()