[Pw_forum] filling the single state (A1g) in DFT+U calculation

[Yusuf Zuntu]

Dear QE Users,

I have calculated U of my model based on linear response approach, but I am yet to understand whether filling up the single state (A1g) as proposed in the tutorials  
http://media.quantum-espresso.org/santa_barbara_2009_07/slides-exercices/Handson_ldau.pdf   

is universal for all systems. I have a system of boron nitride doped with transition metals. Without imposing state on the singlet, the (A1g) tends to have not the lowest state in the minority spin as seen below:
 
relax calculations:

 spin  1
    eigenvalues: 
  0.697  0.758  1.023  1.045  1.048

spin  2
    eigenvalues: 
  0.021  0.023  0.042  0.110  0.118

My question is do i have to fill up the singlet state as recommended in the tutorial or no need. Although I do not have any experimental results to define the say the expected final outcome of the result (whether is semi-conducting, insulator or metallic). I will equally need some physical evidence as to why the imposition of the state on the singlet state. 
Thank you

Yusuf Zuntu
PhD candidate
USM