-
Notifications
You must be signed in to change notification settings - Fork 0
Data Exploration and Analysis
GIS Data Agent v14.5 — 基于 Google ADK 的 AI 地理空间平台 仓库:https://github.com/zhouning/gisdataagent
上传数据后,使用 describe_geodataframe 获取完整的数据质量概要:
- 坐标参考系 (CRS):识别并验证投影信息
- 空值分析:每个字段的空值数量与比例
- 重复检测:几何和属性层面的重复要素统计
- 几何问题:无效几何、自相交、空几何等
- 统计摘要:数值字段的均值、中位数、标准差、分位数
示例提示词:
请描述一下这个数据集的质量情况
Describe the quality profile of this dataset
| 工具 | 功能 | 示例提示词 |
|---|---|---|
| 缓冲区分析 | 生成指定距离的缓冲区 | 对学校点做500米缓冲区 |
| 叠加分析 | 交集、并集、差集运算 | 求两个图层的交集 |
| 裁剪 | 按边界裁剪数据 | 用行政区边界裁剪土地利用数据 |
| 分区统计 | 按区域聚合统计值 | 按街道统计平均房价 |
| 热力图 | 点密度可视化 | 生成犯罪事件热力图 |
| 聚类分析 | 空间聚类(DBSCAN/K-means) | 对POI点进行空间聚类 |
| 邻近分析 | K近邻、距离矩阵 | 找出每个学校最近的3个医院 |
| 方法 | 说明 | 适用场景 |
|---|---|---|
| Moran's I | 全局空间自相关指数 | 判断数据是否存在空间聚集模式 |
| LISA | 局部空间自相关指标 | 识别高-高、低-低、高-低、低-高聚类 |
| Getis-Ord Gi* | 热点/冷点分析 | 发现统计显著的热点和冷点区域 |
示例提示词:
计算房价的Moran's I指数
对人口密度做热点分析
Perform a hotspot analysis on population density
对时空数据进行趋势分析和预测:
预测未来3个月的NDVI变化趋势
Forecast NDVI trends for the next 3 months
模拟参数变化对空间指标的影响:
如果容积率提高20%,对交通拥堵有什么影响?
What if we increase building density by 20%?
中心性分析、可达性评估:
计算道路网络的介数中心性
Analyze road network betweenness centrality
评估设施服务覆盖范围:
分析医院的15分钟步行可达范围
Analyze 15-minute walking accessibility to hospitals
| 功能 | 说明 | 示例提示词 |
|---|---|---|
| NDVI | 归一化植被指数 | 计算这幅影像的NDVI |
| DEM 分析 | 数字高程模型:坡度、坡向、山影 | 分析DEM的坡度分布 |
| LULC 分类 | 土地利用/土地覆盖分类 | 对遥感影像进行土地利用分类 |
使用 MaskablePPO(来自 sb3_contrib)进行土地利用布局优化:
-
自定义环境:
ParcelScoringPolicy评估用地方案 - 优化目标:通过配对农田↔林地置换最小化碎片化,同时保持面积平衡
- 5 种优化场景:支持不同约束条件下的用地规划
-
模型权重:
scorer_weights_v7.pt预训练权重
示例提示词:
优化这个区域的土地利用布局
Optimize the land use layout for this area
10 种融合策略,支持跨格式、跨来源的数据整合:
| 策略 | 说明 |
|---|---|
spatial_join |
基于空间关系的连接 |
attribute_join |
基于属性字段的连接 |
zonal_stats |
分区统计融合 |
overlay_union |
叠加并集 |
overlay_intersection |
叠加交集 |
nearest_join |
最近邻连接 |
temporal_join |
时间窗口连接 |
hierarchical_merge |
层级聚合融合 |
feature_enrichment |
要素属性增强 |
conflict_resolution |
冲突消解融合 |
融合执行支持 PostGIS 下推 加速大规模空间运算。
示例提示词:
将人口数据与行政区划数据基于空间位置融合
Fuse population data with admin boundaries using spatial join
GIS Data Agent v14.5 — AI geospatial platform on Google ADK Repository: https://github.com/zhouning/gisdataagent
After uploading data, use describe_geodataframe to get a full quality profile:
- CRS (Coordinate Reference System): Identify and validate projection info
- Null analysis: Null count and ratio for each field
- Duplicate detection: Duplicate feature statistics at geometry and attribute levels
- Geometry issues: Invalid geometry, self-intersections, empty geometry, etc.
- Statistical summary: Mean, median, standard deviation, and quantiles for numeric fields
Example prompts:
请描述一下这个数据集的质量情况
Describe the quality profile of this dataset
| Tool | Function | Example Prompt |
|---|---|---|
| Buffer | Generate buffers at a specified distance | Create a 500m buffer around schools |
| Overlay | Intersection, union, difference operations | Find the intersection of these two layers |
| Clip | Clip data by boundary | Clip land use data by admin boundary |
| Zonal statistics | Aggregate statistics by zone | Calculate average housing price by district |
| Heatmap | Point density visualization | Generate a crime incident heatmap |
| Clustering | Spatial clustering (DBSCAN/K-means) | Cluster POI points spatially |
| Nearest neighbors | KNN, distance matrix | Find the 3 nearest hospitals for each school |
| Method | Description | Use Case |
|---|---|---|
| Moran's I | Global spatial autocorrelation index | Determine if data exhibits spatial clustering patterns |
| LISA | Local Indicators of Spatial Association | Identify High-High, Low-Low, High-Low, Low-High clusters |
| Getis-Ord Gi* | Hotspot/Coldspot analysis | Find statistically significant hot and cold spots |
Example prompts:
计算房价的Moran's I指数
Perform a hotspot analysis on population density
Perform trend analysis and forecasting on spatiotemporal data:
预测未来3个月的NDVI变化趋势
Forecast NDVI trends for the next 3 months
Simulate the impact of parameter changes on spatial metrics:
如果容积率提高20%,对交通拥堵有什么影响?
What if we increase building density by 20%?
Centrality analysis, accessibility assessment:
计算道路网络的介数中心性
Analyze road network betweenness centrality
Evaluate facility service coverage:
分析医院的15分钟步行可达范围
Analyze 15-minute walking accessibility to hospitals
| Feature | Description | Example Prompt |
|---|---|---|
| NDVI | Normalized Difference Vegetation Index | Calculate NDVI for this image |
| DEM analysis | Digital Elevation Model: slope, aspect, hillshade | Analyze slope distribution from DEM |
| LULC classification | Land Use / Land Cover classification | Classify land use from remote sensing imagery |
Land-use layout optimization using MaskablePPO (from sb3_contrib):
-
Custom environment:
ParcelScoringPolicyevaluates land-use plans - Optimization goal: Minimize fragmentation via paired farmland-forest swaps while maintaining area balance
- 5 optimization scenarios: Support different constraint conditions for land planning
-
Model weights: Pre-trained weights from
scorer_weights_v7.pt
Example prompts:
优化这个区域的土地利用布局
Optimize the land use layout for this area
10 fusion strategies supporting cross-format, cross-source data integration:
| Strategy | Description |
|---|---|
spatial_join |
Join based on spatial relationships |
attribute_join |
Join based on attribute fields |
zonal_stats |
Zonal statistics fusion |
overlay_union |
Overlay union |
overlay_intersection |
Overlay intersection |
nearest_join |
Nearest neighbor join |
temporal_join |
Temporal window join |
hierarchical_merge |
Hierarchical aggregation fusion |
feature_enrichment |
Feature attribute enrichment |
conflict_resolution |
Conflict resolution fusion |
Fusion execution supports PostGIS push-down for accelerating large-scale spatial operations.
Example prompts:
将人口数据与行政区划数据基于空间位置融合
Fuse population data with admin boundaries using spatial join
GIS Data Agent v14.5 · Built on Google ADK · GitHub · Report Issue
快速开始 Getting Started
核心概念 Core Concepts
使用指南 User Guide
架构与运维 Architecture & Ops
参考 Reference