Optimization of lattice with MATTER element

[pratik]

Hi,

I am looking into using the MATTER element with thicknesses less than 0.001
X0. Is there a way to use this element and optimize the lattice?

The probabilistic model does not allow me to easily do this. However, if I can stop it from changing after some evaluations, it could be possible.

Let me know if this could be possible.

elegant.lte

(867 Bytes) Downloaded 1119 times

elegant.ele

(1.5 KiB) Downloaded 1135 times

Best,
Pratik

[michael_borland]

Pratik,

You can try using the &transmute_elements command to turn the MATTER elements into drifts.

Code: Select all

&transmute_elements
        name = *, type = MATTER, new_type = DRIF
&end

&run_setup
...

--Michael

[pratik]

Hi Michael,

Thanks for your reply. I do want the MATTER element to be there. Do you think setting restart_random_numbers = 0, is a better choice. I do get the optimization to work with that change but it is still not monotonically decreasing. I am also switching to parallel_optimization for faster convergence. Let me know what you think about this and if there is any better way to do it. I am attaching the files.

[pratik]

Hi Michael,

Thanks for your reply. I do want the MATTER element to be there. Do you think setting restart_random_numbers = 0, is a better choice. I do get the optimization to work with that change but it is still not monotonically decreasing. I am also switching to parallel_optimization for faster convergence. Let me know what you think about this and if there is any better way to do it. I am attaching the files.

[michael_borland]

If you want to optimize with the MATTER element in place, then setting restart_random_numbers=1 in the optimization_setup command is the right thing to do. That should ensure that the random effects are the same each time.

--Michael