Adding geospatial data to Parquet. Follows the GeoParquet v0.4 spec.
Reading geoparquet files can be done with read.
julia> import GeoParquet as GP
julia> url = "https://github.com/opengeospatial/geoparquet/raw/v0.4.0/examples/example.parquet"
julia> fn = download(url)
julia> df = GP.read(fn)
5×6 DataFrame
Row │ pop_est continent name iso_a3 gdp_md_est geometry
│ Int64? String? String? String? Float64? WellKnow…
─────┼──────────────────────────────────────────────────────────────────────────────────────────────────────────────
1 │ 920938 Oceania Fiji FJI 8374.0 WellKnownBinary{Geom, Vector{UIn…
2 │ 53950935 Africa Tanzania TZA 150600.0 WellKnownBinary{Geom, Vector{UIn…
3 │ 603253 Africa W. Sahara ESH 906.5 WellKnownBinary{Geom, Vector{UIn…
4 │ 35623680 North America Canada CAN 1.674e6 WellKnownBinary{Geom, Vector{UIn…
5 │ 326625791 North America United States of America USA 1.856e7 WellKnownBinary{Geom, Vector{UIn…Writing requires Table like input with geometry columns that are WellKnownBinary from GeoFormatTypes.
julia> GeoParquet.write("test.parquet", df, (:geometry,))
test.parquet- Strict schema and version support
- GeoInterface FeatureCollection implementation
- Better access to metadata in
read - Lazy read option
- Easier support for ProjJSON
We thank Julia Computing for supporting contributions to this package.