tippecanoe

You can control the visibility and organization of individual features within a tileset by adding a tippecanoe object directly to the GeoJSON Feature. This object takes precedence over CLI flags like --layer.

Supported keys in the tippecanoe object:

• minzoom: The minimum zoom level at which the feature is included. Specifying this prevents the feature from being dropped by dot-dropping (-r).
• maxzoom: The maximum zoom level at which the feature is included.
• layer: The name of the layer this feature belongs to.

Note: The tippecanoe object must be a sibling to properties and geometry within the Feature object, not inside properties.

You can run shell commands to filter or transform features as tiles are being assembled using -C (prefilter) and -c (postfilter).

Filter Mechanics:

• Features are passed via stdin as newline-delimited GeoJSON objects.
• The filter must write the resulting GeoJSON features to stdout.
• The shell command receives three arguments: $1 (zoom), $2 (X), and $3 (Y).

Filter Types:

• Prefilter (-C): Runs at highest resolution, before simplification, topology repair, or dropping. It receives features with the tippecanoe element (which may include index, sequence, extent, and dropped). You must pass the tippecanoe element through to maintain internal operations.
• Postfilter (-c): Runs at tile resolution, after simplification and cleaning.

If you are unsure which options to use, start with this command to automatically select a maximum zoom level and drop the least visible features to keep tile sizes manageable:

tippecanoe -zg -o out.mbtiles --drop-densest-as-needed in.geojson

• -zg: Automatically chooses a maximum zoom level sufficient for the data's precision.
• --drop-densest-as-needed: Drops the least visible features at each zoom level to prevent tiles from becoming too large.

You can build and run a tippecanoe environment using Docker to avoid local dependency issues. The following commands build the image from the current directory and run a conversion task, mounting a local directory to /data inside the container.

$ docker build -t tippecanoe:latest .
$ docker run -it --rm \
  -v /tiledata:/data \
  tippecanoe:latest \
  tippecanoe --output=/data/output.mbtiles /data/example.geojson

To build tippecanoe from source on Linux, you must install gcc, g++, make, libsqlite3-dev, and zlib1g-dev.

Note: Tippecanoe requires C++11 or newer. If your default g++ is older, you may need to install a newer version (e.g., g++-5) and set the CXX environment variable.

You can control feature visibility and layer naming directly within your source GeoJSON by adding a tippecanoe object to each feature. This object must be a sibling to properties and geometry (not inside properties).

Supported keys in the tippecanoe object:

• minzoom: The minimum zoom level at which the feature is included.
• maxzoom: The maximum zoom level at which the feature is included.
• layer: The name of the layer this feature belongs to (overrides --layer).

If your source data has these values inside properties, use ndjson-cli to move them into the tippecanoe object before running tippecanoe.

If you are unsure which options to use, start with this pattern to ensure a good balance between detail and tile size:

$ tippecanoe -zg -o out.mbtiles --drop-densest-as-needed in.geojson

• -zg: Automatically chooses a maximum zoom level high enough to reflect the precision of the original data.
• --drop-densest-as-needed: If tiles are too large, Tippecanoe drops the least visible features at each zoom level to keep tile sizes manageable.

When generating tiles, you can control how features are simplified or dropped across different zoom levels using the following parameters:

• --maxzoom (-z): The highest zoom level in the tile set.
• --minzoom (-Z): The lowest zoom level in the tile set.
• --basezoom (-B): The zoom level at which feature dropping begins.
• --drop-rate (-r): The rate at which features are dropped. Note that -r requires either -zg (guess max zoom) or --smallest-maximum-zoom-guess to function correctly.

Example of setting a max zoom and a drop rate:

tippecanoe -z 14 -r 0.5 input.json

Tippecanoe can be installed on macOS using Homebrew or built from source on Ubuntu.

macOS (Homebrew):

$ brew install tippecanoe

Ubuntu (Source):

$ git clone https://github.com/felt/tippecanoe.git
$ cd tippecanoe
$ make -j
$ make install