search

mumax / 3

GPU-accelerated micromagnetic simulator
Other
447 stars 150 forks source link

Feature/angles #232

Closed peytondmurray closed 5 years ago

peytondmurray commented 5 years ago

I needed some functions which give the magnetization in terms of the usual spherical coordinates θ and ϕ. I also needed to know the in-plane component of the magnetization rxy, so I implemented all three as a single vector field ext_rxyphitheta with components (rxy, ϕ, θ). The field is calculated on the GPU before being transferred back to host memory, so it's pretty quick. Would anyone else find this useful?

JeroenMulkers commented 5 years ago

Some users might find this an interesting feature. However, I am resistant to merge the current implementation due to two different reasons.

  1. It does not make much sense to put the three variables in a vectorfield quantity. This will lead to a non-intuitive visualization of this 'vectorfield' in the GUI or when using mumax3-convert.

  2. I do not see the point for having both the in-plane component rxy and the angle theta since rxy=sin theta.

In my opinion it is better to define two scalar fields for the magnetization angles phi and theta, and to not define the in-plane magnetization component rxy.

peytondmurray commented 5 years ago

I agree. I removed the vector field and implemented the theta and phi scalar fields.

godsic commented 5 years ago

@peytondmurray Looks good to me. Please

The ext_ thing is a way to highlight contributed functionality.

peytondmurray commented 5 years ago

Okay, I've attached the ext_ prefix to the scalar fields and renamed engine/angles.go to engine/ext_angles.go. Here's the result of running ./test.bash &> test_result.txt.

godsic commented 5 years ago

Cheers!