skmap.io.base.read_auth_rasters#

read_auth_rasters(raster_files, username, password, bands=None, dtype='float16', n_jobs=4, return_base_raster=False, nodata=None, verbose=False)[source]#

Read raster files trough a authenticate HTTP service, aggregating them into a single array. For raster files without authentication it’s better to use read_rasters.

The nodata value is replaced by np.nan in case of dtype=float*, and for dtype=*int* it’s replaced by the the lowest possible value inside the range (for int16 this value is -32768).

Parameters:

  • raster_files (List) – A list with the raster urls.

  • username (str) – Username to provide to the basic access authentication.

  • password (str) – Password to provide to the basic access authentication.

  • bands – Which bands needs to be read. By default is None reading all the bands.

  • dtype (str) – Convert the read data to specific dtype. By default it reads in float16 to save memory, however pay attention in the precision limitations for this dtype [1].

  • n_jobs (int) – Number of parallel jobs used to read the raster files.

  • return_base_raster (bool) – Return an empty raster with the same properties of the read rasters (height, width, n_bands, crs, dtype, transform).

  • nodata – Use this value if the nodata property is not defined in the read rasters.

  • verbose (bool) – Use True to print the reading progress.

Returns:

A 4D array, where the first dimension refers to the bands and the last dimension to read files. If return_base_raster=True the second value will be a base raster path.

Return type:

Numpy.array or Tuple[Numpy.array, Path]

References

[1] Float16 Precision

Examples

>>> from skmap.io.base import read_auth_rasters
>>>
>>> # Do the registration in
>>> # https://glad.umd.edu/ard/user-registration
>>> username = '<YOUR_USERNAME>'
>>> password = '<YOUR_PASSWORD>'
>>> raster_files = [
...     'https://glad.umd.edu/dataset/landsat_v1.1/47N/092W_47N/850.tif',
...     'https://glad.umd.edu/dataset/landsat_v1.1/47N/092W_47N/851.tif',
...     'https://glad.umd.edu/dataset/landsat_v1.1/47N/092W_47N/852.tif',
...     'https://glad.umd.edu/dataset/landsat_v1.1/47N/092W_47N/853.tif'
... ]
>>>
>>> data, base_raster = read_auth_rasters(
...     raster_files,
...     username,
...     password,
...     return_base_raster=True,
...     verbose=True
... )
>>> print(f'Data: shape={data.shape}, dtype={data.dtype} and base_raster={base_raster}')