Turf-MCP

# 空间聚合和聚类函数 from fastmcp import FastMCP from turf_mcp.utils import call_js_script aggregation_mcp = FastMCP("aggregation") @aggregation_mcp.tool async def collect(polygons: str, points: str, in_field: str, out_field: str) -> str: """ 将点属性聚合到多边形中。 此功能将位于多边形内部的点的属性值聚合到多边形中，用于统计和汇总分析。 Args: polygons: 多边形特征集合 - 类型: str (JSON 字符串格式的 GeoJSON FeatureCollection) - 格式: FeatureCollection with Polygon or MultiPolygon features - 示例: '{"type": "FeatureCollection", "features": [{"type": "Feature", "geometry": {"type": "Polygon", "coordinates": [[[125, -15], [113, -22], [154, -27], [144, -15], [125, -15]]]}}, ...]}' 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], "properties": {"population": 100}}, ...]}' in_field: 输入字段名 - 类型: str - 描述: 点特征中要聚合的属性字段名 - 示例: 'population' out_field: 输出字段名 - 类型: str - 描述: 多边形特征中要创建的聚合属性字段名 - 示例: 'total_population' Returns: str: JSON 字符串格式的 GeoJSON FeatureCollection - 类型: GeoJSON FeatureCollection with Polygon features - 格式: {"type": "FeatureCollection", "features": [{"type": "Feature", "geometry": {"type": "Polygon", "coordinates": [...]}, "properties": {..., "total_population": 聚合值}}, ...]} - 示例: '{"type": "FeatureCollection", "features": [{"type": "Feature", "geometry": {"type": "Polygon", "coordinates": [...]}, "properties": {"total_population": 1500}}, ...]}' 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 >>> polygons = '{"type": "FeatureCollection", "features": [{"type": "Feature", "geometry": {"type": "Polygon", "coordinates": [[[125, -15], [113, -22], [154, -27], [144, -15], [125, -15]]]}}]}' >>> points = '{"type": "FeatureCollection", "features": [{"type": "Feature", "geometry": {"type": "Point", "coordinates": [-75.343, 39.984], "properties": {"population": 100}}]}' >>> result = asyncio.run(collect(polygons, points, 'population', 'total_population')) >>> print(result) '{"type": "FeatureCollection", "features": [{"type": "Feature", "geometry": {"type": "Polygon", "coordinates": [...]}, "properties": {"total_population": 100}}, ...]}' Notes: - 输入参数 polygons 和 points 必须是有效的 JSON 字符串 - 坐标顺序为 [经度, 纬度] (WGS84 坐标系) - 仅对位于多边形内部的点进行属性聚合 - 聚合方式为求和，即计算多边形内部所有点的属性值总和 - 依赖于 Turf.js 库和 Node.js 环境 """ js_script = f""" const turf = require('@turf/turf'); const polygons = JSON.parse('{polygons}'); const points = JSON.parse('{points}'); const inField = '{in_field}'; const outField = '{out_field}'; const result = turf.collect(polygons, points, inField, outField); console.log(JSON.stringify(result)); """ try: return await call_js_script(js_script) except Exception as e: raise Exception(f"执行异常: {str(e)}") @aggregation_mcp.tool async def clustersDbscan(points: str, max_distance: float, options: str = None) -> str: """ 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 """ options_param = options if options else '{}' js_script = f""" const turf = require('@turf/turf'); const points = JSON.parse('{points}'); const maxDistance = parseFloat({max_distance}); const options = JSON.parse('{options_param}'); const result = turf.clustersDbscan(points, maxDistance, options); console.log(JSON.stringify(result)); """ try: return await call_js_script(js_script) except Exception as e: raise Exception(f"执行异常: {str(e)}") @aggregation_mcp.tool async def clustersKmeans(points: str, number_of_clusters: int, options: str = None) -> str: """ 使用 K-means 算法进行点聚类。 此功能使用 K-means 聚类算法对点进行聚类，将点划分为指定数量的簇。 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]}}, ...]}' number_of_clusters: 聚类数量 - 类型: int - 描述: 要创建的聚类数量 - 示例: 5 options: 可选参数配置 - 类型: str (JSON 字符串) 或 None - 可选字段: - numberOfClusters: 聚类数量（与 number_of_clusters 参数相同） - mutate: 是否修改原始特征 (默认: false) - 示例: '{"mutate": true}' 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(clustersKmeans(points, 5)) >>> 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) - K-means 算法需要预先指定聚类数量 - 聚类编号从 0 开始 - 算法使用随机初始中心点，每次运行结果可能不同 - 依赖于 Turf.js 库和 Node.js 环境 """ options_param = options if options else '{}' js_script = f""" const turf = require('@turf/turf'); const points = JSON.parse('{points}'); const numberOfClusters = parseInt({number_of_clusters}); const options = JSON.parse('{options_param}'); const result = turf.clustersKmeans(points, numberOfClusters, options); console.log(JSON.stringify(result)); """ try: return await call_js_script(js_script) except Exception as e: raise Exception(f"执行异常: {str(e)}")