paddle. nanmedian ( x, axis=None, keepdim=False, name=None ) [source]

Compute the median along the specified axis, while ignoring NaNs.

If the valid count of elements is a even number, the average value of both elements in the middle is calculated as the median.

  • x (Tensor) – The input Tensor, it’s data type can be int32, int64, float16, bfloat16, float32, float64.

  • axis (None|int|list|tuple, optional) – The axis along which to perform median calculations axis should be int or list of int. axis should be in range [-D, D), where D is the dimensions of x . If axis is less than 0, it works the same way as \(axis + D\). If axis is None, median is calculated over all elements of x. Default is None.

  • keepdim (bool, optional) – Whether to reserve the reduced dimension(s) in the output Tensor. If keepdim is True, the dimensions of the output Tensor is the same as x except in the reduced dimensions(it is of size 1 in this case). Otherwise, the shape of the output Tensor is squeezed in axis . Default is False.

  • name (str, optional) – Name for the operation (optional, default is None). For more information, please refer to Name.


Tensor, results of median along axis of x. The output dtype is the same as x.


>>> import paddle
>>> x = paddle.to_tensor([[float('nan'), 2. , 3. ], [0. , 1. , 2. ]])

>>> y1 = x.nanmedian()
>>> print(y1.numpy())

>>> y2 = x.nanmedian(0)
>>> print(y2.numpy())
[0.  1.5 2.5]

>>> y3 = x.nanmedian(0, keepdim=True)
>>> print(y3.numpy())
[[0.  1.5 2.5]]

>>> y4 = x.nanmedian((0, 1))
>>> print(y4.numpy())