std

→Tensor

std(input: Tensor, dim: _int | Sequence[_int] | None = None, unbiased: _bool | None = None, keepdim: _bool = False, correction: _int | _float | None = None)

source edit

Implementing kernel

C++std_opfree fn

Compute the standard deviation along dim.

The square root of var with the same arguments: correction=1 (the default) divides by $n-1$ , correction=0 by $n$ .

Parameters

inputTensor

Input tensor.

dimint or sequence of int= None

Dimension(s) to reduce. None reduces every element. A dimension named twice is refused.

unbiasedbool= None

True is correction=1, False correction=0; it overrides correction when given. A bool in the place of dim — std(x, False) — is read as this.

keepdimbool= False

Whether to retain reduced dims with size 1.

correctionint or float= None

Degrees-of-freedom correction. Default 1.

Returns

Tensor

Reduced floating-point tensor.

Raises

ValueError

If dim names one dimension twice.

Notes

\sigma \;=\; \sqrt{ \frac{1}{\max(n - c, 0)} \sum_{i=1}^{n} (x_i - \bar{x})^2 } .

The gradient of a slice with no spread (std 0) is 0, the subgradient the reference picks, rather than nan. Too few elements ( $n \le c$ ) give nan or inf, as var explains.

Examples

>>> import lucid
>>> x = lucid.tensor([1.0, 2.0, 3.0, 4.0])
>>> lucid.std(x)
tensor(1.291)
>>> lucid.std(x, correction=0)
tensor(1.118)

std