Same inputs but different output - seed issue

Hello everyone,
To begin, I want to convey my sincere appreciation for your willingness to assist.

I’m currently working on a space planning project, using the “Termite Nest” plugin in Grasshopper. However, I’ve encountered a persistent issue that seems to affect other graph-based plugins like MagnetizingFPG as well.

The problem I’m facing is that the plugin generates non-unique output layouts, even when provided with identical input variables. It appears that the seed value is continuously changing in the background, which makes it extremely challenging to establish correlations between layouts and their fitness values for optimization purposes.

To highlight the main concern, it primarily revolves around the Spatial-graph component. For instance, when I modify the location or even the index of the second attractor point, the plugin produces varying layouts despite having the same input parameters.

I suspect that this issue might be related to the plugin itself or perhaps other similar plugins in Grasshopper. Unfortunately, I don’t have access to the scripts to investigate further. I’ve even considered the possibility of GHpython being involved, although I did attempt to reinstall Grasshopper without success.

This issue is causing a significant obstacle in my project, as it prevents me from integrating these space planning plugins effectively with multi-objective optimization plugins. Without the ability to retrieve consistent genotypes for identical genes, the optimization process becomes impractical.

I would greatly appreciate any insights, suggestions, or solutions that the community can provide to help resolve this issue. Your assistance would be invaluable to my thesis project.

Thank you in advance for your time and expertise.

Best regards,



3

Relations.csv (182 Bytes)
Space planning.gh (94.0 KB)

Hi,

if its not your code, you always rely on someone else.:man_shrugging: Using many plugins might help you on short terms, but as soon as you run into a limitation you have a big problem. I know that writing a thesis is time consuming, but it shouldn’t be such a big task to write your own spatial network component . Later in life you cannot excuse yourself that its not working, you need to deliver something. A good engineer will always find a solution in time. And the key of doing so is to gain control of what you do in appropriate portions. Clicking together code snippets might solve it, but its no skill, nor does it work most of the times.

1 Like