Locking Genes from a Previous Simulation

Grasshopper Wallacei
1

Hi everyone,

I’m setting up a simulation and want to lock specific genes from a previous run and only allow the remaining other genes to evolve. Is this somehow possible in Wallacei? Is there a best practice workflow?

Thanks!

2

If you lock a gene it means that, it is not a gene. if you don’t want a number parameter to change during simulation just don’t connect it to Wallacei as a gene.

3

Sorry, I wasn’t clear earlier. I have a system that optimizes four objectives, but one of them is more important than the others. The remaining three objectives are only relevant if the first objective reaches a high level of quality.

My plan is to first run a simulation optimizing the primary objective. Since I know which variables (sliders) impact this objective, I’d like to narrow down the list of variables related to those sliders for the next simulation.

I’m considering extracting the genome from the first simulation in Wallacei, replacing the sliders influencing the first objective with a genome list from which a new slider will select variables. Then, I would run a new simulation to optimize objectives 2-4.

Does this make sense?

4

Agreed 100% - You can do this in two ways:

  1. run the algorithm for one objective, identify the fittest solution from the simulation (gen no. and ind. no.) and select that solution’s genome from Wallacei X output and decode the genome (using the decode genome component)… once you decode the genome, you will have the slider values for each gene for that solution, which then you can manually set via non-wallacei sliders in preparation for the second simulation. If you will choose this option, a helpful tip would be to alphabetically name your wallacei sliders/genepools… (for example: wlc_a, wlc_b, wlc_c) and then make sure you connect the sliders/genepools to wallacei by right clicking the wallacei component and selecting “select all sliders”. If you do this, when you decode the selected phenotype’s genome, the slider values will be listed according to their alphabetical order. This will make things a bit easier for you when manually setting slider values.

  2. A faster option, once you identify your optimal phenotype from the first simulation… open the WallaceiX user interface and export just that solution (make sure you are only exporting one solution, and not a group of solutions). After the export completes, the slider values in your grasshopper definition will belong to that exported solution… this way, you do not need to decode genome,… you’ll just need to identify the wallacei sliders/genepools you want to disconnect from the second simulation and disconnect them from the wallacei X component (make sure you remove the wlc_ prefix from these sliders/genepools before starting the second simulation).

Best of luck! would love to see the results once you’re done with the simulation(s)!