Makes a number of modifications to RFISensitivityMask:
First generate a static frequency mask by looking for outlier
channels in the quantile over time.
Iteratively determine both the baseline and the standard deviation about
that baseline by applying weighted rolling medians that accounts for the
current masks.
Remove the baseline before applying the sumthreshold method.
Use the frequency and time dependent standard deviation in the sumthreshold method.
Increase the default maximum window size used in sumthreshold.
Modify the sumthreshold method to allow for:
Full control over how the threshold changes with window size.
Use the expected standard deviation times a constant value for the
sensitivity-based masking.
Calculate the expected noise in the convolution by propagating estimates
of the variance for the individual samples.
Makes a number of modifications to RFISensitivityMask: