Generative design creating single design output error

I started a new post since the issue now is different - Following with our exchange with @jacob.small
The issue I have is that the generative design component is not iterating through the different variables and thus, the output is a single design solution as opposed to several.

  • When manually changing the sliders, the script runs fine. I even checked ALL the values for each slider to make sure I didn’t have any error.
  • I am testing all this in Dynamo Sandbox. No issues found in the script.
    -I deleted all the nodes that I was using that obtained information from the Revit API, I am relying only on remember nodes as input data.

Anyone has any ideas on what else I can try? Maybe its a simple fix and I am not seeing it.
Thanks in advance

@jacob.small
are you trying to minimize values?:

Yes

And if so, are there possible slider combinations which return a null in Dynamo Sandbox?

No. The sliders are not interacting with each other and thus I was able to check all the design options one by one and got no error. Note: I simplified the design space to be able to do so.

Okay, another thing that I tried:
Changing the design to Randomize.
Now, its working.
The issue is apparently when it tries to minimize, as @jacob.small had suggested.
Randomize study works for my purposes, but I would like to know why minimize is not working.

If you plug the exact same inputs which produce a null in GD into the graph manually and run it in Dynamo for Revit or Dynamo Sandbox , do you get results?

I don’t know how to select null results in GD since I am only getting a single design output, and it is not null (it produces a value in the output).
The randomize study is not facing the same issue which is very interesting.

Select one of the optimize runs which only produced null and note the input values used to produce that result.

Open the graph in Dynamo for Revit or Dynamo Sandbox, and then set your inputs to the values which GD used.

There is only a single run (which is the issue, actually). That single run, is not creating any null results.

The examples you previously posted here should still be valid, unless the graph has changed.

Hi @jacob.small
The issue is only when using GD in “optimize” mode. “Randomize” works fine. Here you can see a side by side comparison. Optimize only generates a single option. Randomize generates lots of options and I can open them in Dynamo.
Any clue as to why that might be?
Thanks for your help.


Elena – when you ask to optimize only ONE Output, you get only ONE answer!! If you want to compare more options, optimize for more outputs that compete against each other.

(I realized the experience is lacking here and we should do a better job of conveying this fact!)

That makes sense. Thanks for clarifying. I guess I was expecting to find several “optimal” but that would only be if I have more than one goal.