scipy.ndimage.

median#

scipy.ndimage.median(input, labels=None, index=None)[source]#

Calculate the median of the values of an array over labeled regions.

Parameters:

inputarray_like: Array_like of values. For each region specified by labels, the median value of input over the region is computed.
labelsarray_like, optional: An array_like of integers marking different regions over which the median value of input is to be computed. labels must have the same shape as input. If labels is not specified, the median over the whole array is returned.
indexarray_like, optional: A list of region labels that are taken into account for computing the medians. If index is None, the median over all elements where labels is non-zero is returned.

Returns:

medianfloat or list of floats: List of medians of input over the regions determined by labels and whose index is in index. If index or labels are not specified, a float is returned: the median value of input if labels is None, and the median value of elements where labels is greater than zero if index is None.

See also

label, minimum, maximum, extrema, sum, mean, variance, standard_deviation

Notes

The function returns a Python list and not a NumPy array, use np.array to convert the list to an array.

Array API Standard Support

median has experimental support for Python Array API Standard compatible backends in addition to NumPy. Please consider testing these features by setting an environment variable SCIPY_ARRAY_API=1 and providing CuPy, PyTorch, JAX, or Dask arrays as array arguments. The following combinations of backend and device (or other capability) are supported.

Library	CPU	GPU
NumPy	✅	n/a
CuPy	n/a	✅
PyTorch	✅	⛔
JAX	⚠️ no JIT	⛔
Dask	⚠️ computes graph	n/a

See Support for the array API standard for more information.

Examples

>>> from scipy import ndimage
>>> import numpy as np
>>> a = np.array([[1, 2, 0, 1],
...               [5, 3, 0, 4],
...               [0, 0, 0, 7],
...               [9, 3, 0, 0]])
>>> labels, labels_nb = ndimage.label(a)
>>> labels
array([[1, 1, 0, 2],
       [1, 1, 0, 2],
       [0, 0, 0, 2],
       [3, 3, 0, 0]], dtype=int32)
>>> ndimage.median(a, labels=labels, index=np.arange(1, labels_nb + 1))
[2.5, 4.0, 6.0]
>>> ndimage.median(a)
1.0
>>> ndimage.median(a, labels=labels)
3.0