How to compute the A/B compartment from a real imaging data?

I wonder if the document at present only works for real Hi-C. How can I explicitly compute A/B compartment from a real 
imaging data. 

Plus, can you give a more detailed description/documentation of your igm and MaxComp tools... 
The reviewer asks us to reproduce it and compare with this method, but I find the tutorial is not that 
user friendly and I struggle a lot.... 