wxy command reference
Generated from the help text of wxy 0.0.1, built from wxylib/core@16a14c1. Each section is exactly what wxy <command> --help prints. Worked examples are in the how-tos.
wxy
wxylib from the shell
Usage: wxy <COMMAND>
Commands:
warp Warp a GeoTIFF or Zarr array onto a target grid, written as GeoZarr
points Read a GeoTIFF or Zarr array at points (x,y lines on stdin), each chunk fetched once; values go to stdout as CSV
zonal Summarise a GeoTIFF or Zarr array over GeoJSON polygons (the cells GDAL's rasterizer would burn); statistics go to stdout as CSV
help Print this message or the help of the given subcommand(s)
Options:
-h, --help Print help
-V, --version Print version
wxy warp
Warp a GeoTIFF or Zarr array onto a target grid, written as GeoZarr
Usage: wxy warp [OPTIONS] --crs <CRS> --extent <XMIN> <YMIN> <XMAX> <YMAX> --size <NCOL> <NROW> <SOURCE> <OUTPUT>
Arguments:
<SOURCE> Source: a GeoTIFF file or https:// URL, or a Zarr store directory
<OUTPUT> Output Zarr store directory, created if missing
Options:
--crs <CRS>
Target CRS as an EPSG:NNNN code, such as EPSG:3031. (The CRS backend also reads PROJ strings, but GeoZarr's proj: has no place to record one; WKT2 and PROJJSON wait on a PROJ-backed CRS.)
--extent <XMIN> <YMIN> <XMAX> <YMAX>
Target extent: xmin ymin xmax ymax, in target CRS units
--size <NCOL> <NROW>
Target size in cells: ncol nrow
-r, --resampling <RESAMPLING>
Resampling method [default: near] [possible values: near, bilinear, cubic]
--error-threshold <ERROR_THRESHOLD>
Maximum error of the approximate transformer, in pixels; 0 is exact [default: 0.125]
--nodata <NODATA>
Nodata, overriding the source's
--source-array <SOURCE_ARRAY>
Array to read when the source is Zarr [default: /data]
--array <ARRAY>
Array to write in the output store [default: /data]
--no-wrap
Leave target cells beyond a 360 degree lon/lat source's edge empty, as GDAL's warper does, instead of wrapping longitude and reading pole rows
--plain
Warp in one pass with a 1:1 kernel instead of chunk by chunk as GDAL's warper does. Faster on large targets, but bilinear and cubic then sample rather than smooth a target coarser than the source
--overview <OVERVIEW>
Overview to read from a GeoTIFF with overviews, as gdalwarp's -ovr: auto (the coarsest still finer than the target), auto-N (N levels finer than that), N (overview N, 0 being the first) or none [default: auto]
--bounds
Place a curvilinear Zarr source by its CF cell bounds (the cell corners) instead of by its cell centres as GDAL does
--cache <DIR>
Keep the source's bytes, as fetched, in this cache directory (append-only Parquet fragments) and read from it what it already holds. GeoTIFF sources (files or URLs; URL rows are validated by the server's etag) and byte-reference tables
--remap <TARGET>
For a byte-reference table source, the `remap` target its source paths are rewritten by (such as a mirror, or local copies)
--wm <WM>
Working memory one warp chunk may take, in MB, before it is split into smaller chunks (gdalwarp's -wm) [default: 64]
--overwrite
Replace the output array if it exists
--dry-run
Print the plan as JSON (the chunks it fetches, the bytes that costs, how it is chunked) and stop, reading no data
-h, --help
Print help
wxy points
Read a GeoTIFF or Zarr array at points (x,y lines on stdin), each chunk fetched once; values go to stdout as CSV
Usage: wxy points [OPTIONS] <SOURCE>
Arguments:
<SOURCE> Source: a GeoTIFF file or https:// URL, or a Zarr store directory
Options:
--crs <CRS> CRS of the points (EPSG:NNNN or a PROJ string) [default: EPSG:4326]
--source-array <SOURCE_ARRAY> Array to read in a Zarr source [default: /data]
--bounds Place a curvilinear Zarr source by its CF cell bounds
--no-wrap Leave points beyond a 360 degree lon/lat source's edge off it, as GDAL does, instead of wrapping their longitude
-r, --resampling <RESAMPLING> near reads the cell a point falls in (gdallocationinfo); the others interpolate as GDAL's InterpolateAtPoint, empty where it has no value [default: near] [possible values: near, bilinear, cubic, cubicspline]
--dry-run Print the plan as JSON (the chunks it fetches, the bytes that costs, how it is chunked) and stop, reading no data
-h, --help Print help
wxy zonal
Summarise a GeoTIFF or Zarr array over GeoJSON polygons (the cells GDAL's rasterizer would burn); statistics go to stdout as CSV
Usage: wxy zonal [OPTIONS] <SOURCE> <POLYGONS>
Arguments:
<SOURCE> Source: a GeoTIFF file or https:// URL, or a Zarr store directory
<POLYGONS> GeoJSON: a FeatureCollection, a Feature, or a bare geometry, of Polygon or MultiPolygon
Options:
--crs <CRS> CRS of the polygons (GeoJSON's is lon/lat) [default: EPSG:4326]
--all-touched Every cell a polygon touches, not only those whose centre it covers
--coverage Weight each cell by the fraction of it a polygon covers (exact coverage, as exactextract): count is the covered cells' sum of fractions, sum the fraction-weighted sum
--source-array <SOURCE_ARRAY> Array to read in a Zarr source [default: /data]
--bounds Place a curvilinear Zarr source by its CF cell bounds
--dry-run Print the plan as JSON (the chunks it fetches, the bytes that costs, how it is chunked) and stop, reading no data
-h, --help Print help