Fits with Fitting

[tennant]

Hi,

I've never been very successful in using the optimizer in elegant. I'm attaching a simple lattice file with a request/challenge to get the fit to converge. I have four 90-degree FODO cells. I simply want to move the phase advance (both planes) and maintain the same Twiss parameters at the exit. If this exercise is done with DIMAD, the fit converges and is done in less than a fraction of a second. With elegant, the way I currently have the optimizer configured, it takes a good part of a minute and does not converge. I'm fitting 6 parameters (Bx, By, ax, ay, psix, psiy) with 9 available knobs. I'd love for someone to show me what I'm doing wrong!

Thanks,
chris

[michael_borland]

Chris,

Please try the attached file. I think the main problem was the huge step sizes.

-Michael