rolling_window_sigma_clip (function) ∞
-
rolling_window_sigma_clip
(log, array, clippingSigma, windowSize)[source] ∞ given a sorted list of values, median sigma-clip values based on a window of values either side of each value (rolling window) and return the array mask
Key Arguments:
log
– loggerarray
– the array to clean up (arrays of length < 5 are not clipped but returned unmasked)clippingSigma
– the minimum sigma to clip (using median-absolute distribution as sigma proxy)windowSize
– the size of the window to use when calculating the median distribution (window of 11 will use 5 values each side of the value in question)
Usage:
from fundamentals.stats import rolling_window_sigma_clip arrayMask = rolling_window_sigma_clip( log=self.log, array=myArray, clippingSigma=2.2, windowSize=11) ## JUST KEEP UNMASKED VALUES try: myArray = [e for e, m in zip( myArray, arrayMask) if m == False] except: myArray = []