A question about elements_from_sequence in parttree

[mayy17]

I see

node = part_tree.parent_node((seq_model.gaps[-1], seq_model.z_dir[-1]))

in elements_from_sequence() in parttree
But it seems that z_dir refers to dir of mirrors instead of gaps, so len(z_dir)=len(gaps)+1.
In my optic system with five mirrors, the z_dir of Img surf is +1(maybe in default?), but z_dir[-2]=-1, and actually the z_dir of airgap is -1 . It causes the tree cannot find the node.
Maybe it should be

node = part_tree.parent_node((seq_model.gaps[-1], seq_model.z_dir[-2]))
?

[mjhoptics]

Hello @mayy17,

Thank you for the report. The intent of the code is that gaps, z_dir, and rndx are all the same length, and ifcs, lcl_tfrms, and gbl_tfrms are the same and longer by 1. Having said that, it hasn't always been that way and I can imagine there are some code paths that produce lists that don't conform to this intent.

I tried several models but wasn't able to duplicate the problem. It would be helpful if:

1. clarify which version of ray-optics you're using
2. you could post the model (or a dumbed down version if sensitive) and/or script that produces the error
3. If item 2 isn't possible, please clarify how the model was generated, i.e. Zemax or CODE V file, .roa file, direct model construction via script (using either add_surface or the opm methods), or ...

Thank you for using ray-optics.
Regards,
Mike Hayford

[mayy17]

So thanks for your reply and your works.
I create my model in zemax 19.4 and try to read it with rayoptics==0.8.5. And I also tried code in https://ray-optics.readthedocs.io/en/latest/_modules/rayoptics/zemax/zmxread.html#read_lens_file but got the same result.
I've attached a simplified model which may help. I succeed to duplicate the problem with it.
LENS.zip

And with the previous code I can read the model and get result of raytrace, maybe this problem is just caused by a bug which creates a wrong model in read_lens_file() in zmxread.

[mjhoptics]

Hi @mayy17,
I just updated a fix to the master branch; the postprocessing of the zemax file import needed tweaking. The commit is f055ded.
HTH
Mike

[mayy17]

It works well. Thanks for your help.