Open geejaytee opened 6 years ago
As CMB is primarily used for positive scores, the plots assume that everything is >0 (which is true once rescaling is performed). If the data contains negative values, then the plots disappear (as the lower bound in ylim for the individual plots is fixed at 0)
geejaytee, thanks for trying the CMB and posting comments. As you mention in the very beginning, the program was primarily designed to analyse a symmetrically distributed around the center nucleosome positioning data. The idea of the algorithm, of cause, could be applied to any matrices but in current version the support is limited The very unfortunate is that I don't have anymore access to a MatLab and can't implement any changes to the program myself but, of cause, will be happy if somebody will make the program better :) In my plans was to convert the CMB clustering tool to a R package or at least R script some time later, but so far I don't have spare time. Thanks for your contributions! Definitely a real user cases is very important, because when I am using the program myself I am avoiding most of the bugs just because I am using it only in particular way.
Just started using this to cluster some (non-nucleosome) data, but I've come up with some issues: