Aller au contenu

Preprocessing data

Mapbox vector tiles from GeoJSON

The easiest way to use tippecanoe on Windows is through WSL.

Guess the max level:

tippecanoe -zg -e [dest dir] -pC --drop-densest-as-needed --extend-zooms-if-still-dropping [input.geojson]

Specify the max level:

tippecanoe -z[max level] -e [dest dir] -pC --drop-densest-as-needed --extend-zooms-if-still-dropping [input.geojson]

Extracting OSM vector data

We have a big pmtiles file with all the OSM data, but it’s too big!

In order to extract it into smaller datasets, we have a tool, https://gitlab.siradel.com/horizon/tools/extract-mvt, composed of 3 scripts.

  • dl_osm.py, downloads a subset of the tiles pyramid.
  • cull.py, filters the tiles to keep only some layers, features, and attributes.
  • to_pmtiles.py, repackages the result into a pmtiles file.

Note that none of this is production-ready, generic, or configurable. So good luck using it. This paragraph might even be obsolete.

DTM from tiff raster

Use gdal2tiles from https://gitlab.siradel.com/tpetillon/gdal2tiles and texture_converter from https://gitlab.siradel.com/siliciumserver/TextureConverter.

gdal2tiles.py -f tiff -z [minzoom]-[maxzoom] -r cubic [input file] [output folder]
texture_converter -f tiff -t float -i [input folder] -o [output folder] -n [nodata value]

For the tiling step, the -z parameter can be omitted to let the tool decide on the best minimum and maximum levels to generate. Also the maxzoom can be omitted alone. For example one may ask the tool to generate tiles from level 0 to whatever max level it sees fit with -z 0-.

Using the raster tiler

Documentation

Upload file

scp <file> connect@lfrn1ubuntu10:/shared_data/depot/<file>

Submit computation

POST http://lfrn1ubuntu10:18213/rasters-tiler/compute

{
    "inputFiles": [
        "/shared_data/depot/<file>"
    ],
    "tmsProfile": "global-mercator",
    "type": "single_band_real"
}

Get status

GET http://lfrn1ubuntu10:18213/rasters-tiler/{id}/status

Download result

scp -r connect@lfrn1ubuntu10:/shared_data/rasters-tiler/results/{id} result

Delete computation

DELETE http://lfrn1ubuntu10:18213/rasters-tiler/{id}