1. Can a Parquet reader open the table?
viewparquet can answer this first question by showing rows, column types, and raw stored values. It is a useful intake check, but a successful read is not GeoParquet validation.
GeoParquet
GeoParquet viewer in your browser
GeoParquet stores geospatial vector data — points, lines, and polygons — in Apache Parquet, with a specification for identifying geometry columns and related geospatial metadata.
viewparquet opens GeoParquet through its Parquet reader. You can browse the attribute table, inspect column names and DuckDB types, see geometry values in their stored form, filter non-spatial attributes with SQL, and export results. It does not currently render geometries on a map, decode them into GIS features, or validate GeoParquet CRS and metadata fields; use a geo-aware tool when you need those checks.
Updated August 21, 2026 · Reproducible checks, limitations, and primary references are included below.
Inspect a GeoParquet tableInspect a GeoParquet file
Start with the upstream five-row example
Download the GeoParquet project's current five-row example from the verified public-datasets guide linked below, then drag its .parquet file into viewparquet. We rechecked that upstream file with the shipped DuckDB-WASM v1.4.3 reader on August 19, 2026; the date matters because a main-branch example can change.
Compare the six-column schema
In that check, the reader exposed pop_est (DOUBLE), continent, name, and iso_a3 (VARCHAR), gdp_md_est (BIGINT), and geometry (BLOB). This confirms how the current table reader exposes this particular file; it does not prove that another file has valid GeoParquet metadata or CRS information.
Reproduce the attribute check
Run SELECT count(*) AS rows, count(DISTINCT continent) AS continents FROM data. The checked example returned 5 rows and 3 continents. Attribute SQL and tabular exports work without implying map rendering, geometry decoding, spatial predicates, CRS checks, or GeoParquet validation.
Separate three different GeoParquet checks
A file can be readable as Parquet while its geospatial metadata is incomplete or its geometries are invalid. Treat table access, specification conformance, and GIS validity as separate questions.
2. Does the file follow a GeoParquet version?
A geo-aware validator must inspect the required file metadata, primary geometry column, encoding, geometry types, CRS declaration, and version-specific rules. GeoParquet 1.x and the 2.0 release-candidate line should not be described as one interchangeable schema.
3. Are the geometries and CRS meaningful?
Validity, expected bounds, axis order, and spatial relationships require geometry decoding and domain checks. Use a GIS or spatial engine for this layer; the current viewparquet table does not perform it.
Verify the checked example at the attribute layer
After opening the upstream five-row example described above, run this attribute-only query.
SELECT
count(*) AS rows,
count(DISTINCT continent) AS continents,
count(geometry) AS non_null_geometry
FROM data;What to expect: The copy checked on August 19, 2026 returned 5 rows, 3 continents, and 5 non-null stored geometry values. That still does not prove the geometry encoding, CRS metadata, or shapes are valid.
Primary references
These sources support the format and tool behavior described on this page. Product-specific boundaries are checked against the shipped viewer using the claim-verification method.
Common questions
What is a GeoParquet file?
GeoParquet is Apache Parquet with a standardized way to identify geospatial vector data. It defines geometry columns and geo metadata such as encoding, geometry types, CRS, and optional bounds; a general Parquet reader can read the table, while a geo-aware tool interprets those fields.
How do I open a GeoParquet file without installing QGIS or GDAL?
Open it in viewparquet for a browser-based table and schema inspection. DuckDB-WASM reads a local file without a GIS install, but map rendering, decoded geometry inspection, spatial validation, and CRS checks still require a geo-aware tool.
How is the geometry column stored and displayed?
The current GeoParquet specification uses WKB (Well-Known Binary) carried by Parquet GEOMETRY or GEOGRAPHY logical types. Earlier GeoParquet versions used WKB in BYTE_ARRAY columns and also allowed single-geometry encodings, so older files and readers can expose geometry differently. viewparquet displays the column value and DuckDB type returned by its reader, but does not currently draw the geometry, decode it into an interactive GIS feature, or validate the encoding.
What should I see in the GeoParquet project example?
With the upstream example checked on August 19, 2026, viewparquet's shipped DuckDB-WASM reader returned 5 rows across pop_est, continent, name, iso_a3, gdp_md_est, and a raw geometry column. Use that small file to reproduce table and attribute-SQL behavior, not as proof that viewparquet validates GeoParquet metadata or CRS correctness.
GeoParquet vs shapefile — why switch?
GeoParquet is a compressed columnar file with richer schemas and a single-file layout, while shapefiles split a dataset across sidecar files and impose legacy field constraints. GeoParquet also fits analytical engines and object-storage workflows, but GIS compatibility should still be checked for your tools.
Is my GeoParquet data uploaded when I view it?
A local GeoParquet file is read in your browser with DuckDB-WASM and is not uploaded to viewparquet. A remote object is requested directly by your browser from its host; aggregate analytics never includes the filename, remote URL, geometry values, or file contents.
Continue with a related task
Try it on your own file
Drop a Parquet, GeoParquet, CSV, or JSON file into viewparquet: browse rows, run DuckDB SQL, and export results in-browser without uploading the dataset to viewparquet. Optional AI sends messages and applicable context directly to the provider you choose; review Settings → AI before use.