Turf-MCP

使用 DBSCAN 算法进行点聚类。

此功能使用基于密度的空间聚类算法 (DBSCAN) 对点进行聚类，识别密集区域。

Args: points: 点特征集合 - 类型: str (JSON 字符串格式的 GeoJSON FeatureCollection) - 格式: FeatureCollection with Point features - 示例: '{"type": "FeatureCollection", "features": [{"type": "Feature", "geometry": {"type": "Point", "coordinates": [-75.343, 39.984]}}, ...]}'

max_distance: 最大距离
    - 类型: float
    - 描述: 聚类搜索的最大距离（单位：千米）
    - 示例: 100.0

options: 可选参数配置
    - 类型: str (JSON 字符串) 或 None
    - 可选字段:
        - units: 距离单位 (默认: 'kilometers')
            - 有效值: 'miles', 'nauticalmiles', 'kilometers', 'meters', 'yards', 'feet', 'inches'
        - minPoints: 形成聚类所需的最小点数 (默认: 3)
    - 示例: '{"units": "miles", "minPoints": 5}'

Returns: str: JSON 字符串格式的 GeoJSON FeatureCollection - 类型: GeoJSON FeatureCollection with Point features - 格式: {"type": "FeatureCollection", "features": [{"type": "Feature", "geometry": {"type": "Point", "coordinates": [lng, lat]}, "properties": {"cluster": 聚类编号, ...}}, ...]} - 示例: '{"type": "FeatureCollection", "features": [{"type": "Feature", "geometry": {"type": "Point", "coordinates": [-75.343, 39.984]}, "properties": {"cluster": 1}}, ...]}'

Raises: Exception: 当 JavaScript 执行失败、超时或输入数据格式错误时抛出异常

Example: >>> import asyncio >>> points = '{"type": "FeatureCollection", "features": [{"type": "Feature", "geometry": {"type": "Point", "coordinates": [-75.343, 39.984]}}]}' >>> result = asyncio.run(clustersDbscan(points, 100.0, '{"minPoints": 3}')) >>> print(result) '{"type": "FeatureCollection", "features": [{"type": "Feature", "geometry": {"type": "Point", "coordinates": [-75.343, 39.984]}, "properties": {"cluster": 1}}, ...]}'

Notes: - 输入参数 points 和 options 必须是有效的 JSON 字符串 - 坐标顺序为 [经度, 纬度] (WGS84 坐标系) - DBSCAN 算法能够识别任意形状的聚类，并处理噪声点 - 聚类编号从 0 开始，-1 表示噪声点（不属于任何聚类） - 依赖于 Turf.js 库和 Node.js 环境