geomaster

Name: geomaster
Author: K-Dense-AI

31kstars3.1kforks

How to install geomaster?

k-dense-ai/scientific-agent-skills/geomaster

$npx -y skills add K-Dense-AI/scientific-agent-skills --skill geomaster

Installs into the current project.

Summary

Comprehensive geospatial science skill covering remote sensing, GIS, spatial analysis, machine learning for earth observation, and 30+ scientific domains. Supports satellite imagery processing (Sentinel, Landsat, MODIS, SAR, hyperspectral), vector and raster data operations, spat

Files · 1

View on GitHub

SKILL.md

1	# GeoMaster
2
3	Comprehensive geospatial science skill covering GIS, remote sensing, spatial analysis, and ML for Earth observation across 70+ topics with 500+ code examples in 8 programming languages.
4
5	## Installation
6
7	```bash
8	# Core Python stack (conda recommended)
9	conda install -c conda-forge gdal rasterio fiona shapely pyproj geopandas
10
11	# Remote sensing & ML
12	uv pip install rsgislib torchgeo earthengine-api
13	uv pip install scikit-learn xgboost torch-geometric
14
15	# Network & visualization
16	uv pip install osmnx networkx folium keplergl
17	uv pip install cartopy contextily mapclassify
18
19	# Big data & cloud
20	uv pip install xarray rioxarray dask-geopandas
21	uv pip install pystac-client planetary-computer
22
23	# Point clouds
24	uv pip install laspy pylas open3d pdal
25
26	# Databases
27	conda install -c conda-forge postgis spatialite
28	```
29
30	## Quick Start
31
32	### NDVI from Sentinel-2
33
34	```python
35	import rasterio
36	import numpy as np
37
38	with rasterio.open('sentinel2.tif') as src:
39	red = src.read(4).astype(float) # B04
40	nir = src.read(8).astype(float) # B08
41	ndvi = (nir - red) / (nir + red + 1e-8)
42	ndvi = np.nan_to_num(ndvi, nan=0)
43
44	profile = src.profile
45	profile.update(count=1, dtype=rasterio.float32)
46
47	with rasterio.open('ndvi.tif', 'w', **profile) as dst:
48	dst.write(ndvi.astype(rasterio.float32), 1)
49	```
50
51	### Spatial Analysis with GeoPandas
52
53	```python
54	import geopandas as gpd
55
56	# Load and ensure same CRS
57	zones = gpd.read_file('zones.geojson')
58	points = gpd.read_file('points.geojson')
59
60	if zones.crs != points.crs:
61	points = points.to_crs(zones.crs)
62
63	# Spatial join and statistics
64	joined = gpd.sjoin(points, zones, how='inner', predicate='within')
65	stats = joined.groupby('zone_id').agg({
66	'value': ['count', 'mean', 'std', 'min', 'max']
67	}).round(2)
68	```
69
70	### Google Earth Engine Time Series
71
72	```python
73	import ee
74	import pandas as pd
75
76	ee.Initialize(project='your-project')
77	roi = ee.Geometry.Point([-122.4, 37.7]).buffer(10000)
78
79	s2 = (ee.ImageCollection('COPERNICUS/S2_SR_HARMONIZED')
80	.filterBounds(roi)
81	.filterDate('2020-01-01', '2023-12-31')
82	.filter(ee.Filter.lt('CLOUDY_PIXEL_PERCENTAGE', 20)))
83
84	def add_ndvi(img):
85	return img.addBands(img.normalizedDifference(['B8', 'B4']).rename('NDVI'))
86
87	s2_ndvi = s2.map(add_ndvi)
88
89	def extract_series(image):
90	stats = image.reduceRegion(ee.Reducer.mean(), roi.centroid(), scale=10, maxPixels=1e9)
91	return ee.Feature(None, {'date': image.date().format('YYYY-MM-dd'), 'ndvi': stats.get('NDVI')})
92
93	series = s2_ndvi.map(extract_series).getInfo()
94	df = pd.DataFrame([f['properties'] for f in series['features']])
95	df['date'] = pd.to_datetime(df['date'])
96	```
97
98	## Core Concepts
99
100	### Data Types
101
102	\| Type \| Examples \| Libraries \|
103	\|------\|----------\|-----------\|
104	\| Vector \| Shapefile, GeoJSON, GeoPackage \| GeoPandas, Fiona, GDAL \|
105	\| Raster \| GeoTIFF, NetCDF, COG \| Rasterio, Xarray, GDAL \|
106	\| Point Cloud \| LAS, LAZ \| Laspy, PDAL, Open3D \|
107
108	### Coordinate Systems
109
110	- EPSG:4326 (WGS 84) - Geographic, lat/lon, use for storage
111	- EPSG:3857 (Web Mercator) - Web maps only (don't use for area/distance!)
112	- EPSG:326xx/327xx (UTM) - Metric calculations, <1% distortion per zone
113	- Use `gdf.estimate_utm_crs()` for automatic UTM detection
114
115	```python
116	# Always check CRS before operations
117	assert gdf1.crs == gdf2.crs, "CRS mismatch!"
118
119	# For area/distance calculations, use projected CRS
120	gdf_metric = gdf.to_crs(gdf.estimate_utm_crs())
121	area_sqm = gdf_metric.geometry.area
122	```
123
124	### OGC Standards
125
126	- WMS: Web Map Service - raster maps
127	- WFS: Web Feature Service - vector data
128	- WCS: Web Coverage Service - raster coverage
129	- STAC: Spatiotemporal Asset Catalog - modern metadata
130
131	## Common Operations
132
133	### Spectral Indices
134
135	```python
136	def calculate_indices(image_path):
137	"""NDVI, EVI, SAVI, NDWI from Sentinel-2."""
138	with rasterio.open(image_path) as src:
139	B02, B03, B04, B08, B11 = [src.read(i).astype(float) for i in [1,2,3,4,5]]
140
141	ndvi = (B08 - B04) / (B08 + B04 + 1e-8)
142	evi = 2.5 * (B08 - B04) / (B08 + 6B04 - 7.5B02 + 1)
143	savi = ((B08 - B04) / (B08 + B04 + 0.5)) * 1.5
144	ndwi = (B03 - B08) / (B03 + B08 + 1e-8)
145
146	return {'NDVI': ndvi, 'EVI': evi, 'SA

People who liked this also viewed

6 picks

Type

geomaster

31kstars3.1kforks

View on GitHub

How to install geomaster?

k-dense-ai/scientific-agent-skills/geomaster

$npx -y skills add K-Dense-AI/scientific-agent-skills --skill geomaster

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# GeoMaster
2
3	Comprehensive geospatial science skill covering GIS, remote sensing, spatial analysis, and ML for Earth observation across 70+ topics with 500+ code examples in 8 programming languages.
4
5	## Installation
6
7	```bash
8	# Core Python stack (conda recommended)
9	conda install -c conda-forge gdal rasterio fiona shapely pyproj geopandas
10
11	# Remote sensing & ML
12	uv pip install rsgislib torchgeo earthengine-api
13	uv pip install scikit-learn xgboost torch-geometric
14
15	# Network & visualization
16	uv pip install osmnx networkx folium keplergl
17	uv pip install cartopy contextily mapclassify
18
19	# Big data & cloud
20	uv pip install xarray rioxarray dask-geopandas
21	uv pip install pystac-client planetary-computer
22
23	# Point clouds
24	uv pip install laspy pylas open3d pdal
25
26	# Databases
27	conda install -c conda-forge postgis spatialite
28	```
29
30	## Quick Start
31
32	### NDVI from Sentinel-2
33
34	```python
35	import rasterio
36	import numpy as np
37
38	with rasterio.open('sentinel2.tif') as src:
39	red = src.read(4).astype(float) # B04
40	nir = src.read(8).astype(float) # B08
41	ndvi = (nir - red) / (nir + red + 1e-8)
42	ndvi = np.nan_to_num(ndvi, nan=0)
43
44	profile = src.profile
45	profile.update(count=1, dtype=rasterio.float32)
46
47	with rasterio.open('ndvi.tif', 'w', **profile) as dst:
48	dst.write(ndvi.astype(rasterio.float32), 1)
49	```
50
51	### Spatial Analysis with GeoPandas
52
53	```python
54	import geopandas as gpd
55
56	# Load and ensure same CRS
57	zones = gpd.read_file('zones.geojson')
58	points = gpd.read_file('points.geojson')
59
60	if zones.crs != points.crs:
61	points = points.to_crs(zones.crs)
62
63	# Spatial join and statistics
64	joined = gpd.sjoin(points, zones, how='inner', predicate='within')
65	stats = joined.groupby('zone_id').agg({
66	'value': ['count', 'mean', 'std', 'min', 'max']
67	}).round(2)
68	```
69
70	### Google Earth Engine Time Series
71
72	```python
73	import ee
74	import pandas as pd
75
76	ee.Initialize(project='your-project')
77	roi = ee.Geometry.Point([-122.4, 37.7]).buffer(10000)
78
79	s2 = (ee.ImageCollection('COPERNICUS/S2_SR_HARMONIZED')
80	.filterBounds(roi)
81	.filterDate('2020-01-01', '2023-12-31')
82	.filter(ee.Filter.lt('CLOUDY_PIXEL_PERCENTAGE', 20)))
83
84	def add_ndvi(img):
85	return img.addBands(img.normalizedDifference(['B8', 'B4']).rename('NDVI'))
86
87	s2_ndvi = s2.map(add_ndvi)
88
89	def extract_series(image):
90	stats = image.reduceRegion(ee.Reducer.mean(), roi.centroid(), scale=10, maxPixels=1e9)
91	return ee.Feature(None, {'date': image.date().format('YYYY-MM-dd'), 'ndvi': stats.get('NDVI')})
92
93	series = s2_ndvi.map(extract_series).getInfo()
94	df = pd.DataFrame([f['properties'] for f in series['features']])
95	df['date'] = pd.to_datetime(df['date'])
96	```
97
98	## Core Concepts
99
100	### Data Types
101
102	\| Type \| Examples \| Libraries \|
103	\|------\|----------\|-----------\|
104	\| Vector \| Shapefile, GeoJSON, GeoPackage \| GeoPandas, Fiona, GDAL \|
105	\| Raster \| GeoTIFF, NetCDF, COG \| Rasterio, Xarray, GDAL \|
106	\| Point Cloud \| LAS, LAZ \| Laspy, PDAL, Open3D \|
107
108	### Coordinate Systems
109
110	- EPSG:4326 (WGS 84) - Geographic, lat/lon, use for storage
111	- EPSG:3857 (Web Mercator) - Web maps only (don't use for area/distance!)
112	- EPSG:326xx/327xx (UTM) - Metric calculations, <1% distortion per zone
113	- Use `gdf.estimate_utm_crs()` for automatic UTM detection
114
115	```python
116	# Always check CRS before operations
117	assert gdf1.crs == gdf2.crs, "CRS mismatch!"
118
119	# For area/distance calculations, use projected CRS
120	gdf_metric = gdf.to_crs(gdf.estimate_utm_crs())
121	area_sqm = gdf_metric.geometry.area
122	```
123
124	### OGC Standards
125
126	- WMS: Web Map Service - raster maps
127	- WFS: Web Feature Service - vector data
128	- WCS: Web Coverage Service - raster coverage
129	- STAC: Spatiotemporal Asset Catalog - modern metadata
130
131	## Common Operations
132
133	### Spectral Indices
134
135	```python
136	def calculate_indices(image_path):
137	"""NDVI, EVI, SAVI, NDWI from Sentinel-2."""
138	with rasterio.open(image_path) as src:
139	B02, B03, B04, B08, B11 = [src.read(i).astype(float) for i in [1,2,3,4,5]]
140
141	ndvi = (B08 - B04) / (B08 + B04 + 1e-8)
142	evi = 2.5 * (B08 - B04) / (B08 + 6B04 - 7.5B02 + 1)
143	savi = ((B08 - B04) / (B08 + B04 + 0.5)) * 1.5
144	ndwi = (B03 - B08) / (B03 + B08 + 1e-8)
145
146	return {'NDVI': ndvi, 'EVI': evi, 'SA

People who liked this also viewed

6 picks

Type