mars.tensor.indices(dimensions, dtype=<class 'int'>, chunk_size=None)[source]

Return a tensor representing the indices of a grid.

Compute a tensor where the subtensors contain index values 0,1,… varying only along the corresponding axis.

  • dimensions (sequence of ints) – The shape of the grid.

  • dtype (dtype, optional) – Data type of the result.

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


grid – The tensor of grid indices, grid.shape = (len(dimensions),) + tuple(dimensions).

Return type


See also

mgrid, meshgrid


The output shape is obtained by prepending the number of dimensions in front of the tuple of dimensions, i.e. if dimensions is a tuple (r0, ..., rN-1) of length N, the output shape is (N,r0,...,rN-1).

The subtensors grid[k] contains the N-D array of indices along the k-th axis. Explicitly:

grid[k,i0,i1,...,iN-1] = ik


>>> import mars.tensor as mt
>>> grid = mt.indices((2, 3))
>>> grid.shape
(2, 2, 3)
>>> grid[0].execute()        # row indices
array([[0, 0, 0],
       [1, 1, 1]])
>>> grid[1].execute()        # column indices
array([[0, 1, 2],
       [0, 1, 2]])

The indices can be used as an index into a tensor.

>>> x = mt.arange(20).reshape(5, 4)
>>> row, col = mt.indices((2, 3))
>>> # x[row, col]  # TODO(jisheng): accomplish this if multiple fancy indexing is supported

Note that it would be more straightforward in the above example to extract the required elements directly with x[:2, :3].