a question of QP

[lengfei5]

Dear Capybara team,

Thanks for developing this QP extension for transition cell type annotation.
I just have a basic question regarding QP when I read your paper.
In the identify score vector f_t, there are two constrains,

1. f_t between 0 and 1
2. sum of f_t <=1

I tend to interpret the f_t here as identify fraction (0% - 100%) of each cell identifies in the reference.
It is natural that the sum is equal to 1 or 100%.
But I get a bit confused by the < 1, meaning there are some cell identities missing in the reference ?
Thank you.