trim_mean#
- scipy.stats.trim_mean(a, proportiontocut, axis=0)[source]#
Return mean of array after trimming a specified fraction of extreme values
Removes the specified proportion of elements from each end of the sorted array, then computes the mean of the remaining elements.
- Parameters:
- aarray_like
Input array.
- proportiontocutfloat
Fraction of the most positive and most negative elements to remove. When the specified proportion does not result in an integer number of elements, the number of elements to trim is rounded down.
- axisint or None, default: 0
Axis along which the trimmed means are computed. If None, compute over the raveled array.
- Returns:
- trim_meanndarray
Mean of trimmed array.
See also
Notes
For 1-D array a,
trim_mean
is approximately equivalent to the following calculation:import numpy as np a = np.sort(a) m = int(proportiontocut * len(a)) np.mean(a[m: len(a) - m])
trim_mean
has experimental support for Python Array API Standard compatible backends in addition to NumPy. Please consider testing these features by setting an environment variableSCIPY_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
⛔
⛔
Dask
⛔
n/a
See Support for the array API standard for more information.
Examples
>>> import numpy as np >>> from scipy import stats >>> x = [1, 2, 3, 5] >>> stats.trim_mean(x, 0.25) 2.5
When the specified proportion does not result in an integer number of elements, the number of elements to trim is rounded down.
>>> stats.trim_mean(x, 0.24999) == np.mean(x) True
Use axis to specify the axis along which the calculation is performed.
>>> x2 = [[1, 2, 3, 5], ... [10, 20, 30, 50]] >>> stats.trim_mean(x2, 0.25) array([ 5.5, 11. , 16.5, 27.5]) >>> stats.trim_mean(x2, 0.25, axis=1) array([ 2.5, 25. ])