Rasters Sources Using GDAL

Dataset.open_raster(key, path, driver='GTiff', options=(), mode='r')[source]

Open a raster file within this Dataset under key. Only metadata are kept in memory.

>>> help(GDALFileRaster)

Parameters

key: hashable (like a string)

File identifier within Dataset

To avoid using a key, you may use aopen_raster()

path: string
driver: string

gdal driver to use when opening the file http://www.gdal.org/formats_list.html

options: sequence of str

options for gdal

mode: one of {‘r’, ‘w’}

Returns

source: GDALFileRaster

Example

>>> ds.open_raster('ortho', '/path/to/ortho.tif')
>>> file_proj4 = ds.ortho.proj4_stored

>>> ds.open_raster('dem', '/path/to/dem.tif', mode='w')
>>> nodata_value = ds.dem.nodata

See Also

Dataset.create_raster(key, path, fp, dtype, channel_count, channels_schema=None, driver='GTiff', options=(), sr=None, ow=False, **kwargs)[source]

Create a raster file and register it under key within this Dataset. Only metadata are kept in memory.

The raster’s values are initialized with channels_schema[‘nodata’] or 0.

>>> help(GDALFileRaster)
>>> help(GDALMemRaster)

Parameters

key: hashable (like a string)

File identifier within Dataset

To avoid using a key, you may use acreate_raster()

path: string

Anything that makes sense to GDAL:

  • A path to a file

  • An empty string when using driver=MEM

  • A path or an xml string when using driver=VRT

fp: Footprint

Description of the location and size of the raster to create.

dtype: numpy type (or any alias)
channel_count: integer

number of channels

channels_schema: dict or None

Channel(s) metadata. (see Channels schema fields below)

driver: string

gdal driver to use when opening the file http://www.gdal.org/formats_list.html

options: sequence of str

options for gdal http://www.gdal.org/frmt_gtiff.html

sr: string or None

Spatial reference of the new file.

In order not to set a spatial reference, use None.

In order to set a spatial reference, use a string that can be converted to WKT by GDAL.

ow: bool

Overwrite. Whether or not to erase the existing files.

Returns

source: GDALFileRaster or GDALMemRaster

The type depends on the driver parameter

Example

>>> ds.create_raster('dem_copy', 'dem_copy.tif', ds.dem.fp, ds.dsm.dtype, len(ds.dem))
>>> array = ds.dem.get_data()
>>> ds.dem_copy.set_data(array)

Channel schema fields

Fields:

‘nodata’: None or number ‘interpretation’: None or str ‘offset’: None or number ‘scale’: None or number ‘mask’: None or str

Interpretation values:

undefined, grayindex, paletteindex, redband, greenband, blueband, alphaband, hueband, saturationband, lightnessband, cyanband, magentaband, yellowband, blackband

Mask values:

all_valid, per_dataset, alpha, nodata

Additionally:

  • A field missing or None is kept to default value.

  • A field can be passed as

    • a value: All bands are set to this value

    • a sequence of values of length channel_count: All bands will be set to their respective state

Caveat

When using the GTiff driver, specifying a mask or interpretation field may lead to unexpected results.

See Also

Dataset.aopen_raster(path, driver='GTiff', options=(), mode='r')[source]

Open a raster file anonymously within this Dataset. Only metadata are kept in memory.

See open_raster()

Example

>>> ortho = ds.aopen_raster('/path/to/ortho.tif')
>>> file_wkt = ortho.wkt_stored

See Also

Dataset.acreate_raster(path, fp, dtype, channel_count, channels_schema=None, driver='GTiff', options=(), sr=None, ow=False, **kwargs)[source]

Create a raster file anonymously within this Dataset. Only metadata are kept in memory.

See create_raster()

Example

>>> mask = ds.acreate_raster('mask.tif', ds.dem.fp, bool, 1, options=['SPARSE_OK=YES'])
>>> open_options = mask.open_options

>>> channels_schema = {
...     'nodata': -32767,
...     'interpretation': ['blackband', 'cyanband'],
... }
>>> out = ds.acreate_raster('output.tif', ds.dem.fp, 'float32', 2, channels_schema)
>>> band_interpretation = out.channels_schema['interpretation']

See Also