SigmaClip.jl finds outliers in numeric arrays with iterative sigma clipping. It returns clipped copies, clips in place, builds validity masks, reports the final clipping bounds, and computes statistics on the retained values. Scratch buffers can be reused across calls, and the center and spread estimators can be replaced with any function. The package has no runtime dependencies.
using Pkg
Pkg.add("SigmaClip")using SigmaClip
data = [0, 1, 2, 3, 4, 5, 6, 50, NaN, Inf]
sigma_clip(data, 3, 3) #== sigma_clip(data, lower, upper), or sigma_clip(data, 3)
# [0.0, 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, NaN, NaN, NaN]
sigma_clip_mask(data, 3) # validity mask: true = finite and retained
# Bool[1, 1, 1, 1, 1, 1, 1, 0, 0, 0]
sigma_clip_bounds(data, 3)
# (-5.89561331103361, 11.89561331103361)
sigma_clipped_stats(data, 3)
# (center = 3.0, spread = 2.9652044370112036)By default the center is the median and the spread is the median absolute deviation scaled to a normal standard deviation, with at most 5 iterations. See the documentation for how the algorithm treats non-finite and excluded values, in-place clipping, custom statistics, reusable workspaces, and the API reference.
Benchmark comparison with Astropy:
SigmaClip.jl is licensed under the MIT License. See LICENSE.