- tensorstore.cast(base: TensorStore, dtype: dtype) TensorStore
Returns a read/write view with the data type converted.
Example
>>> array = ts.array([1.5, 2.5, 3.5, 4.5, 5.5], dtype=ts.float32) >>> view = ts.cast(array, ts.uint32) >>> view TensorStore({ 'base': { 'array': [1.5, 2.5, 3.5, 4.5, 5.5], 'driver': 'array', 'dtype': 'float32', }, 'context': {'data_copy_concurrency': {}}, 'driver': 'cast', 'dtype': 'uint32', 'transform': {'input_exclusive_max': [5], 'input_inclusive_min': [0]}, }) >>> await view.read() array([1, 2, 3, 4, 5], dtype=uint32)