`scipy.stats._stats_py:sigmaclip`

source: /scipy/stats/_stats_py.py:3377

Signature

def sigmaclip(a, low = 4.0, high = 4.0, *, nan_policy = propagate)

Summary

Perform iterative sigma-clipping of array elements.

Extended Summary

Starting from the full sample, all elements outside the critical range are removed, i.e. all elements of the input array c that satisfy either of the following conditions

c < mean(c) - std(c)*low
c > mean(c) + std(c)*high

The iteration continues with the updated sample until no elements are outside the (updated) range.

Parameters

a : array_like

Data array, will be raveled if not 1-D.

low : float, optional

Lower bound factor of sigma clipping. Default is 4.

high : float, optional

Upper bound factor of sigma clipping. Default is 4.

nan_policy : {'propagate', 'omit', 'raise'}

Defines how to handle input NaNs.

propagate: if a NaN is present in the input, the clipped array will be empty, and the upper and lower thresholds will be NaN.
omit: NaNs will be omitted when performing the calculation.
raise: if a NaN is present, a ValueError will be raised.

Returns

clipped : ndarray: Input array with clipped elements removed.
lower : float: Lower threshold value use for clipping.
upper : float: Upper threshold value use for clipping.

Notes

This function iteratively removes observations. Once observations are removed, they are not re-added in subsequent iterations. Consequently, although it is often the case that clipped is identical to a[(a >= lower) & (a <= upper)], this property is not guaranteed to be satisfied; clipped may have fewer elements.

Array API Standard Support

sigmaclip 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             ⚠️ no JIT           
Dask                  ⛔                     n/a                 
====================  ====================  ====================

See dev-arrayapi for more information.

Examples

import numpy as np
from scipy.stats import sigmaclip
a = np.concatenate((np.linspace(9.5, 10.5, 31),
                    np.linspace(0, 20, 5)))
fact = 1.5
c, low, upp = sigmaclip(a, fact, fact)

✓

c
c.var(), c.std()
low, c.mean() - fact*c.std(), c.min()
upp, c.mean() + fact*c.std(), c.max()

✗

(np.float64(10.035355339059327), np.float64(10.035355339059327), np.float64(10.033333333333333))

a = np.concatenate((np.linspace(9.5, 10.5, 11),
                    np.linspace(-100, -50, 3)))
c, low, upp = sigmaclip(a, 1.8, 1.8)

✓

(c == np.linspace(9.5, 10.5, 11)).all()

✗

np.True_

Aliases

scipy.stats.sigmaclip