element#1.Particles in optimization with Pelegant

[wmliu]

Hi,

When I use element#1.Particles to optimize for transmission using elegant, every thing works perfectly. But when I tried to speed up with Pelegant, I'm getting error like
"Problem evaluating expression: MENDARC0#1.Particles MSOL3LTR#1.Particles - MENDARC0#1.Particles /
RPN expression errors prevent optimization
error: division by zero"
It looks like, the information about the number of particles is not available when optimizing with Pelegant. Is there an alternative way of doing with Pelegant?

Thanks,

Wanming

[ywang25]

Wanming,

Can you provide a simplified test case, so we can reproduce the problem?

Thanks,

Yusong

[wmliu]

Yusong,

Please find the attached file.
In this example, because the lattice has been optimized, so elegant will finish very fast. But when I do it with mpirun -n 6 Pelegant ....., it ended up with the error message.
If I commented out the optimization term with number of particles involved, both elegant and Pelegant works the same except Pelegant is way more faster.

Thanks,

Wanming

[wmliu]

Yusong,

I managed to get around it by replace the number of particles with the known number at the beginning and still use the number of particles at the end of the last element.

Thanks,

Wanming

[wmliu]

Yusong,

Sorry for the misleading information in my previous post. The problem is still there. Because the number of particles at elements are missing with Pelegant, the optimization term will always be 1 after I made the previous change. So I only get away with the error message but not the real problem to be solved. I'll just use elegant when ever I need to optimize the transmission for now.

Thanks,

Wanming

[ywang25]

Wanming,

A bug has been found for the simplex method in Pelegant when the optimization term has "transmission". It will be fixed in the next release.

Yusong