mindspore.mint.var_mean

mindspore.mint.var_mean(input, dim=None, *, correction=1, keepdim=False)

By default, return the variance and mean of each dimension in Tensor. If dim is a dimension list, calculate the variance and mean of the corresponding dimension.

The variance (\(\sigma ^2\)) is calculated as:

\[\sigma ^2 = \frac{1}{N - \delta N} \sum_{j=0}^{N-1} \left(self_{ij} - \overline{x_{i}}\right)^{2}\]

where is \(x\) the sample set of elements, \(\bar{x}\) is the sample mean, \(N\) is the number of samples and \(\delta N\) is the correction .

Warning

This is an experimental API that is subject to change or deletion.

Parameters

• input (Tensor) – The input tensor. Supported dtypes: float16, float32.

• dim (Union[int, tuple(int), list(int)], optional) – Specify the dimensions for calculating variance and mean. Default value: None.

Keyword Arguments

• correction (int, optional) – Difference between the sample size and sample degrees of freedom. Defaults to Bessel's correction. Default: 1.

• keepdim (bool, optional) – Whether to preserve the dimensions of the output Tensor. If True, retain the reduced dimension with a size of 1. Otherwise, remove the dimensions. Default value: False.

Returns

A tuple of variance and mean.

Raises

• TypeError – If input is not a Tensor.

• TypeError – If dim is not one of the following data types: int, tuple, list, or Tensor.

• TypeError – If keepdim is not a bool.

• ValueError – If dim is out of range.

Supported Platforms:

Ascend

Examples

>>> import mindspore as ms
>>> input = ms.Tensor([[1, 2, 3, 4], [-1, 1, 4, -10]], ms.float32)
>>> output_var, output_mean = ms.mint.var_mean(input, 1, correction=2, keepdim=True)
>>> print(output_var)
[[ 2.5]
 [54.5]]
>>> print(output_mean)
[[ 2.5]
 [-1.5]]