mindspore.ops.min

mindspore.ops.min(input, axis=None, keepdims=False, *, initial=None, where=None)[source]

Return the minimum values and their indices along the given axis of the tensor.

Parameters

input (Tensor) – The input tensor.
axis (int) – Specify the axis for computation. If None , compute all elements in the input . Default None .
keepdims (bool) – Whether the output tensor has dim retained. Default False .

Keyword Arguments

initial (scalar, optional) – Initial value for the minimum. Default None .
where (Tensor[bool], optional) – Specifies the range over which to compute the maximum values. The shape of this tensor must be broadcastable to the shape of input . An initial value must be specified. Default None , indicating that all elements are to be computed.

Returns

Tuple(min, min_indices) of 2 tensors.

Supported Platforms:: Ascend GPU CPU

Examples

>>> import mindspore
>>> input = mindspore.tensor([[2, 5, 1, 6],
...                           [3, -7, -2, 4],
...                           [8, -4, 1, -3]])
>>> # case 1: By default, compute the minimum of all elements.
>>> mindspore.ops.min(input)
(Tensor(shape=[], dtype=Int64, value= -7),
 Tensor(shape=[], dtype=Int64, value= 0))
>>>
>>> # case 2: Compute minimum along axis 1.
>>> mindspore.ops.min(input, axis=1)
(Tensor(shape=[3], dtype=Int64, value= [ 1, -7, -4]),
 Tensor(shape=[3], dtype=Int64, value= [2, 1, 1]))
>>>
>>> # case 3: If keepdims=True, the output shape will be same of that of the input.
>>> mindspore.ops.min(input, axis=1, keepdims=True)
(Tensor(shape=[3, 1], dtype=Int64, value=
 [[ 1],
  [-7],
  [-4]]),
 Tensor(shape=[3, 1], dtype=Int64, value=
 [[2],
  [1],
  [1]]))
>>>
>>> # case 4: Use "where" to include only specific elements in computing the minimum.
>>> where = mindspore.tensor([[1, 0, 1, 0],
...                           [0, 0, 1, 1],
...                           [1, 1, 1, 0]], dtype=mindspore.bool_)
>>> mindspore.ops.min(input, axis=1, keepdims=True, initial=0, where=where)
(Tensor(shape=[3, 1], dtype=Int64, value=
 [[ 0],
  [-2],
  [-4]]),
 Tensor(shape=[3, 1], dtype=Int64, value=
 [[0],
  [2],
  [1]]))
>>>
>>> # case 5: The shape of "where" must be broadcast compatible with input.
>>> where = mindspore.tensor([[False],
...                           [False],
...                           [False]])
>>> mindspore.ops.min(input, axis=0, keepdims=True, initial=0, where=where)
(Tensor(shape=[1, 4], dtype=Int64, value=
 [[0, 0, 0, 0]]),
 Tensor(shape=[1, 4], dtype=Int64, value=
 [[0, 0, 0, 0]]))