Learning Forum

[wcai05]

The y_brandh_2D.gds file generated by running y_branch_opt_2D.py gives an error when opening.

[Guilin Sun]

The error message says that there exists an object named "input wg". Please name the object with different names and then do the GDS out.

[mustafacc]

I've tried that already (on wcai's behalf) and the error pops again except with the new different name.

[Guilin Sun]

Please check your FDTD file, and make sure no any objects have the same name.

[mustafacc]

Here's what the objects look like

[Guilin Sun]

Does the new name with "output"? if so, please modify them. Next time it is better to post the error message so it can help us to dig out the issue.

[mustafacc]

"Does the new name with output?" can you clarify what you mean by that?
The error is still the same except with the new name of the cell:


I also tried changing the name of the other objects and the error is the same.


Note this GDS error is occurring after running the default example application note here without changing anything: https://support.lumerical.com/hc/en-us/articles/360042305274-Inverse-design-of-y-branch

I also tried reading the file using various GDS viewer tools.

[Guilin Sun]

I guess there is something wrong with the layer definition. Since the "sub" is actually surrounding the waveguide acting as both cladding and substrate, I just output the waveguide section, and it works well, as long as you define the layer zmin and zmax correctly. Here is my rest result:
All 4 parts are output correctly.

Please check your definition about the layer_def. you can add the substrate and cladding, and have 3 layers, by properly defining their z range. Please try again.

[mustafacc]

Here's the sequence of commands i'm using:
I replicated what was scripted on the app. note. It seems identical to what you have above. However, the error with the GDS file above still persists. You can find the final iteration lms file as well as the generated .gds file: https://drive.google.com/drive/folders/1LNzYo_7JVSDfkuHE-3eUZP44Wn1Ya3Ub?usp=sharing

[Guilin Sun]

Do you mean this error message:


It is unlikely that the script works fine without any warning but the error message pops up. If you mean the error message is from GDS file its self when you use 3rd party software to open, unfortunately we could not help since we do not have that file. To verify the output GDS file from Lumerical script, you can open a new project, import the GDS into it and check if you can see anything wrong. Maybe you need to test the whole process in a fresh new folder.

[mustafacc]

Error message: Yes this is the one.
I uploaded the files (.GDS and .lms) to the directory here: https://drive.google.com/drive/folders/1LNzYo_7JVSDfkuHE-3eUZP44Wn1Ya3Ub?usp=sharing
I can see that the output of the Lumerical GDS export script is wrong because every object is duplicated. You can see this by trying to import the generated GDS file into a project:

The standard GDS format does not allow for duplicate cell names and will result in an error

[wcai05]

Hi, the gds generator files in the y-branch inverse design does not work. Currently i am replacing the "Lumerical_GDS_auto_export.lsfx" files from the ones that i found in the "Y branch optimization using particle swarm algorithm" example and run the .lsf gds generator script to get a gds file with the desier shape.

[mar]

This is because of the Klayout version. 0.27 + is not supported. I had the same problem. With 0.26 this works.
Hope this can be of help

[Guilin Sun]

@mar thank you for the help! In Lumerical output, the result is correct so it must be due to the 3rd party tool.

[wcai05]

Hi, My code for 3D inverse design simulation has been stuck at "Initializing super optimization" for days yet no error message or anything new has shown.
I could not find a solution to this problem.
I wonder if anyone can help?
Thank you!

[Guilin Sun]

I guess that you installed your own version of Python. Please uninstall it, and just use the built-in Python software.