mars.tensor.empty(shape, dtype=None, chunk_size=None, gpu=None, order='C')[source]#

Return a new tensor of given shape and type, without initializing entries.

  • shape (int or tuple of int) – Shape of the empty tensor

  • dtype (data-type, optional) – Desired output data-type.

  • chunk_size (int or tuple of int or tuple of ints, optional) – Desired chunk size on each dimension

  • gpu (bool, optional) – Allocate the tensor on GPU if True, False as default

  • order ({'C', 'F'}, optional, default: 'C') – Whether to store multi-dimensional data in row-major (C-style) or column-major (Fortran-style) order in memory.


out – Tensor of uninitialized (arbitrary) data of the given shape, dtype, and order. Object arrays will be initialized to None.

Return type


See also

empty_like, zeros, ones


empty, unlike zeros, does not set the array values to zero, and may therefore be marginally faster. On the other hand, it requires the user to manually set all the values in the array, and should be used with caution.


>>> import mars.tensor as mt
>>> mt.empty([2, 2]).execute()
array([[ -9.74499359e+001,   6.69583040e-309],
       [  2.13182611e-314,   3.06959433e-309]])         #random
>>> mt.empty([2, 2], dtype=int).execute()
array([[-1073741821, -1067949133],
       [  496041986,    19249760]])                     #random