Install bash compeltion for CLI into ~/.bashrc file. Run it once after install.

geo-gremlin download_wms_tiles \
    WMS_SERVER \
    ZOOM_LEVEL \
    BOUNDS_FN \
    OUTPUT_DIR \
    [--n-workers=100] \
    [--first-n-tiles=N] \
    [--wld-suffix=.wld]

Download map tiles intersecting a GeoJSON boundary. Tiles outside the exact boundary are masked, and a world file is created for every downloaded image. Tiles are loaded in EPSG:3857 projection (web-mercator).

geo-gremlin download_wms_tiles \
    "esri" \
    19 \
    data/test-wms/bounds.geojson \
    ./test

This creates the downloaded tiles in test/esri_19/.

Parameters:

• WMS_SERVER: tile provider: "esri", "google", "bing", "yandex", or "mapbox".
• ZOOM_LEVEL: zoom level used to select tiles, 19 is usually the best you can get.
• BOUNDS_FN: path to a vector file defining the download boundary as Polygon feature.
• OUTPUT_DIR: parent directory for the generated <server>_<zoom> folder.
• --n-workers: number of concurrent download workers. Defaults to 100.
• --first-n-tiles: optionally limit the number of downloaded tiles.
• --wld-suffix: suffix used for generated world files. Defaults to .wld.

geo-gremlin merge-wms-tiles \
    INPUT_DIR \
    [--output=OUTPUT_TIF] \
    [--crs=EPSG:3857]

Merge georeferenced PNG or WebP tiles into a single GeoTIFF.

geo-gremlin merge-wms-tiles test/esri_19/

When --output is omitted, this example creates test/esri_19.tif.

Parameters:

• INPUT_DIR: directory containing image tiles and their world files.
• --output: output GeoTIFF path. By default, a .tif file is created alongside the input directory.
• --crs: coordinate reference system assigned to the output. Defaults to EPSG:3857.

Requires GDAL and its Python bindings.

geo-gremlin retile-wms-tiles \
    INPUT_DIR \
    OUTPUT_DIR \
    [--tile-width=512] \
    [--tile-height=512] \
    [--tile-overlap=0] \
    [--tile-base-name=tile]

Retile georeferenced PNG tiles into uniformly sized tiles. Empty tiles are removed, and partial edge tiles are padded to the requested dimensions.

geo-gremlin retile-wms-tiles \
    test/esri_19/ \
    test/esri_19_retiled/

Parameters:

• INPUT_DIR: directory containing PNG tiles and their world files.
• OUTPUT_DIR: directory where the retiled PNG files will be created.
• --tile-width: output tile width in pixels. Defaults to 512.
• --tile-height: output tile height in pixels. Defaults to 512.
• --tile-overlap: overlap between adjacent tiles in pixels. Defaults to 0.
• --tile-base-name: base name used for generated tiles. Defaults to tile.

Requires GDAL and its Python bindings.

orthogonalize(
    gdf: gpd.GeoDataFrame,
    *,
    n_workers: int = 6,
    verbose: bool = False,
) -> gpd.GeoDataFrame

Orthogonalize all polygon geometries in a GeoDataFrame. It is a batch wrapper over orthogonalize_poly, with optional multiprocessing.

from geo_gremlin.vector.ortho import orthogonalize

ortho_gdf = orthogonalize(overlap_gdf, n_workers=1, verbose=False)

Parameters:

• gdf: gpd.GeoDataFrame, input GeoDataFrame with polygon geometries.
• n_workers: int, number of worker processes (1 runs sequentially).
• verbose: bool, show progress bars when True.

resolve_overlaps(
    gdf: gpd.GeoDataFrame,
    mode: ResolveMode = "union",
    verbose: bool = False,
    **params,
) -> gpd.GeoDataFrame

Resolve intersecting polygons in a GeoDataFrame by grouping overlaps and merging them using a selected mode.

from geo_gremlin.vector.overlap import resolve_overlaps

overlap_gdf = resolve_overlaps(test_gdf, mode="union", verbose=False)

Parameters:

• gdf: gpd.GeoDataFrame, input GeoDataFrame with polygon geometries.
• mode: ResolveMode, merge strategy: "union", "intersection", "largest", or "iou_select".
• verbose: bool, show progress bars when True.
• **params: extra params for mode-specific behavior (iou_threshold for "iou_select").