Compute the variance along the specified axis, while ignoring NaNs. Returns the variance of the array elements, a measure of the spread of a distribution. The variance is computed for the flattened array by default, otherwise over the specified axis. For all-NaN slices or sli
(a, axis=None, dtype=None, out=None, ddof=0, keepdims=np._NoValue,
*, where=np._NoValue)
| 1616 | |
| 1617 | @array_function_dispatch(_nanvar_dispatcher) |
| 1618 | def nanvar(a, axis=None, dtype=None, out=None, ddof=0, keepdims=np._NoValue, |
| 1619 | *, where=np._NoValue): |
| 1620 | """ |
| 1621 | Compute the variance along the specified axis, while ignoring NaNs. |
| 1622 | |
| 1623 | Returns the variance of the array elements, a measure of the spread of |
| 1624 | a distribution. The variance is computed for the flattened array by |
| 1625 | default, otherwise over the specified axis. |
| 1626 | |
| 1627 | For all-NaN slices or slices with zero degrees of freedom, NaN is |
| 1628 | returned and a `RuntimeWarning` is raised. |
| 1629 | |
| 1630 | .. versionadded:: 1.8.0 |
| 1631 | |
| 1632 | Parameters |
| 1633 | ---------- |
| 1634 | a : array_like |
| 1635 | Array containing numbers whose variance is desired. If `a` is not an |
| 1636 | array, a conversion is attempted. |
| 1637 | axis : {int, tuple of int, None}, optional |
| 1638 | Axis or axes along which the variance is computed. The default is to compute |
| 1639 | the variance of the flattened array. |
| 1640 | dtype : data-type, optional |
| 1641 | Type to use in computing the variance. For arrays of integer type |
| 1642 | the default is `float64`; for arrays of float types it is the same as |
| 1643 | the array type. |
| 1644 | out : ndarray, optional |
| 1645 | Alternate output array in which to place the result. It must have |
| 1646 | the same shape as the expected output, but the type is cast if |
| 1647 | necessary. |
| 1648 | ddof : int, optional |
| 1649 | "Delta Degrees of Freedom": the divisor used in the calculation is |
| 1650 | ``N - ddof``, where ``N`` represents the number of non-NaN |
| 1651 | elements. By default `ddof` is zero. |
| 1652 | keepdims : bool, optional |
| 1653 | If this is set to True, the axes which are reduced are left |
| 1654 | in the result as dimensions with size one. With this option, |
| 1655 | the result will broadcast correctly against the original `a`. |
| 1656 | where : array_like of bool, optional |
| 1657 | Elements to include in the variance. See `~numpy.ufunc.reduce` for |
| 1658 | details. |
| 1659 | |
| 1660 | .. versionadded:: 1.22.0 |
| 1661 | |
| 1662 | Returns |
| 1663 | ------- |
| 1664 | variance : ndarray, see dtype parameter above |
| 1665 | If `out` is None, return a new array containing the variance, |
| 1666 | otherwise return a reference to the output array. If ddof is >= the |
| 1667 | number of non-NaN elements in a slice or the slice contains only |
| 1668 | NaNs, then the result for that slice is NaN. |
| 1669 | |
| 1670 | See Also |
| 1671 | -------- |
| 1672 | std : Standard deviation |
| 1673 | mean : Average |
| 1674 | var : Variance while not ignoring NaNs |
| 1675 | nanstd, nanmean |