methylcheck.assign(df, two_pass=False)
Manually and interactively assign each sample to a group, based on beta-value distribution shape.
This function is not documented or supported anymore.
sorts samples by the position of their peak intensity in beta dist, then lets the human assign a number to each. assumes that peaks that differ by > 0.1 of the 0 to 1 range are different clusters. fills in expected cluster #.
if two_pass==True: it lets user go through every cluster a second time and split larger clusters further. sometimes a cluster isn’t obviously two groups until it is separated from the larger dataset.