Closed jbarnoud closed 8 years ago
It's good comment. The idea behind _1 and _2 was to insist on the fact that you to need to know how to assign PBs before performing any stats. Here is my proposal :
Since the clustering part is often slow, it could be nice to separate statistics from the clustering. What do you think?
Looks good to me.
Demo1 and demo2 are not explicit names. There is no way to know what are the demos about without reading or running them.
I suggest to rename demo1 into demo_assignation, and demo2 into demo_analysis. I do not know if the names are the best suitable, though.