Metadata-Version: 2.4
Name: gorkha
Version: 0.0.1
Summary: GORKHA: Graph-based Operational Reconnaissance Kit for Human-in-the-loop Assessment. Post-earthquake damage estimation and survey planning from sparse field surveys.
Author-email: Angshuman Deb <angshuman311@gmail.com>
License-Expression: MIT
Project-URL: Homepage, https://github.com/angshuman311/gorkha
Project-URL: Issues, https://github.com/angshuman311/gorkha/issues
Keywords: earthquake,damage assessment,graph neural network,kriging,survey planning,Nepal
Classifier: Development Status :: 2 - Pre-Alpha
Classifier: Intended Audience :: Science/Research
Classifier: Programming Language :: Python :: 3
Classifier: Topic :: Scientific/Engineering :: GIS
Requires-Python: >=3.10
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: geopandas
Requires-Dist: shapely
Requires-Dist: pyogrio
Requires-Dist: rasterio
Requires-Dist: rasterstats
Requires-Dist: pandas
Requires-Dist: pyarrow
Requires-Dist: numpy
Requires-Dist: scikit-learn
Requires-Dist: scipy
Requires-Dist: pykrige
Requires-Dist: libpysal
Requires-Dist: esda
Requires-Dist: matplotlib
Requires-Dist: requests
Requires-Dist: rapidfuzz
Requires-Dist: joblib
Provides-Extra: deep
Requires-Dist: torch; extra == "deep"
Requires-Dist: torch_geometric; extra == "deep"
Requires-Dist: torchgeo; extra == "deep"
Provides-Extra: sar
Requires-Dist: asf_search; extra == "sar"
Requires-Dist: hyp3_sdk; extra == "sar"
Provides-Extra: test
Requires-Dist: pytest; extra == "test"
Dynamic: license-file

# GORKHA

**Graph-based Operational Reconnaissance Kit for Human-in-the-loop Assessment**

GORKHA estimates building damage in all wards of an earthquake region from a small number
of field surveys, and tells the survey teams where to go next. It uses a graph of the wards,
the early data that are available without a visit (ShakeMap, terrain, satellite images), and
the road network with its blockages.

The test case is the 2015 Gorkha earthquake sequence in Nepal (945 wards, 11 districts).

Status: research code for a CS230 course project. The interfaces change without notice.

## Install

```bash
pip install gorkha
```

For the deep learning parts: `pip install "gorkha[deep]"`. For the Sentinel-1 processing at
the Alaska Satellite Facility: `pip install "gorkha[sar]"`.

## Data

The scripts download public data: the Nepal damage survey (NPC and Kathmandu Living Labs),
OpenStreetMap boundaries and roads (ODbL), USGS ShakeMaps, the NASA ARIA Damage Proxy Map,
the USGS landslide inventory, the Copernicus DEM, Sentinel-1 and Sentinel-2 images. The
Sentinel-1 coherence needs a free NASA Earthdata account.

## License

MIT. See LICENSE. The data sets have their own licenses (see the data section).
