... With a stack of input images, one can identify bad pixels by comparing the deviation of a pixel from the mean with a user­specified threshold times the clipping sigma. ... By uniform weighting, we mean that the weight is a constant for all pixels in an input image. ... In the code, the trimmed mean or median is used. ... If the error maps exist for input images, one does not have to compute the clipping sigma, because the sigma is available from the input error maps for each individual pixels. ...