greenwich 0.7.1

Adding Python conveniences to the wonderful world of GDAL.

Greenwich provides a wrapper for the GDAL SWIG Python bindings. The focus here is on providing some higher level behavior mainly to the raster side of the GDAL/OGR fence.

Installation

The GDAL Python bindings need a little assistance in finding the pertinent headers when building from within a virtualenv. If the usual pip install greenwich fails, specify the GDAL header directory wherever that may be for you such as:

CFLAGS=-I/usr/include/gdal pip install greenwich

Basic Usage

Open any raster data set you have lying around, perhaps some climate data from WorldClim.

from greenwich import Geometry, Raster
from greenwich.io import MemFileIO

with Raster('cc85tn701.tif') as tmax:
    # Save as a NetCDF file.
    tmax.save('cc85tn701.nc')
    geom = Geometry(
        wkt='POLYGON((-123 47,-123 48,-122 49,-121 48,-121 47,-123 47))',
        srs=4326)
    # Clip the raster with a geometry and save the result as a GeoTIFF.
    with tmax.clip(geom) as clipped:
        clipped.save('clipped.tif')

    # Return a NumPy MaskedArray using no data values for a given bounding box.
    m = tmax.masked_array((-120, 38, -118, 44))

    # Convert to an Erdas Imagine file in memory.
    imgio = MemFileIO(suffix='.img')
    tmax.save(imgio)
    imgdata = imgio.read()
    imgio.close()

    # Iterate over bands and retrieve the maximum pixel values.
    maxvals = [band.GetMaximum() for band in tmax]

Retrieve a NumPy array for a specific area by providing the extent as a 4-tuple of min/max x, y coordinates:

arr = tmax.array((-120, 38, -118, 44))

Reproject the raster to another coordinate system. You may pass EPSG codes, WKT, proj4 formatted projections, or a SpatialReference instance as an argument.:

warped = tmax.warp(3857)

Perhaps you would like to resample your image to a new resolution which can be achieved with:

resampled = tmax.resample((100, 100))

Raster instances still behave like a gdal.Dataset.:

meta = tmax.GetMetadata()