rasterio.dtypes module

Mapping of GDAL to Numpy data types.

Since 0.13 we are not importing numpy here and data types are strings. Happily strings can be used throughout Numpy and so existing code will not break.

Within Rasterio, to test data types, we use Numpy’s dtype() factory to do something like this:

if np.dtype(destination.dtype) == np.dtype(rasterio.uint8): …

rasterio.dtypes.can_cast_dtype(values, dtype)

Test if values can be cast to dtype without loss of information.

Parameters
  • values (list-like) –

  • dtype (numpy dtype or string) –

Returns

True if values can be cast to data type.

Return type

boolean

rasterio.dtypes.check_dtype(dt)

Check if dtype is a known dtype.

rasterio.dtypes.get_minimum_dtype(values)

Determine minimum type to represent values.

Uses range checking to determine the minimum integer or floating point data type required to represent values.

Parameters

values (list-like) –

Returns

Return type

rasterio dtype string

rasterio.dtypes.is_ndarray(array)

Check if array is a ndarray.

rasterio.dtypes.validate_dtype(values, valid_dtypes)

Test if dtype of values is one of valid_dtypes.

Parameters
  • values (list-like) –

  • valid_dtypes (list-like) – list of valid dtype strings, e.g., (‘int16’, ‘int32’)

Returns

True if dtype of values is one of valid_dtypes

Return type

boolean