A building is simulated well when I use daylight and energy analysis in honeybee. However, when the colibri plugins is added, the simulation goes on well at early hours, but after several hours Rhino occurs to crash.
@0oxianxiano0, Have you tried to use the
Not sure the issue is related to using Colibri. Probably you need to check the variables, sliders’ value and possible combination to ensure that there isn’t a problem during the iterations.
I’ve experienced this issue as well and I’ve usually found that Grasshopper becomes ustable after consecutively running ~1,000 energy simulations.
I had an exchange with David Rutten about it and, given that it only seems to happen after 1,000 simulations, we agreed that this is not an unreasonable limitation of a desktop application.
I do have a way around this issue for energy simulations, though. If you just use colibri to write out the IDFs (each with a different filename) and record all of the exported idf file paths, you can use the “Rerun IDF” component to run all of the IDFs together after you have used colibri to export them all. With this method, I have not experienced Grasshopper instabilities and I’ve been able to generate and run a few thousand IDFs withlut Grasshopper crashing.
Another benefit of this method is that you can is the “Rerun IDF” parallel capability to run each IDF on a different core, which significantly cuts down calculation time.
The only clunky aspect of this workflow is that you have to make a final script after all of this to read in the IDF results and produce the Colibri CSV but this isn’t too difficult for more advanced GH users.
So this is the workflow that I would recommend if you need to run a sensitivity study with more than 1000 energy simulations.
Interesting! I rarely encountered this issue, (before I switched to
Rerun method), even more than 1000 simulations (with medium size and complexity).
Thank you @chris.
Good to know @OmidmRashidi . Maybe it has to do with stability on certain systems. The computers where I keep encountering this issue were using Windows 7 and 8. We’re yours running Windows 10 ?
Yes @chris , I was run on Window 10.
Thanks for confirming, @OmidmRashidi .
That’s great to know and I’m looking forward to trying this again on a Windows 10 computer.
I’m doing a parametic study with Colibri and honeybee and I’m perform the procedure @chris explained with OSM files instead of IDFs(first generating OSM files and then runing all the simulations with "Re-Run OSM"component). but my file crashes after 400-600 iterations when I try to generate OSM files.
I tried with IDF and it was fine, the reason I’m working with OSM files, is that apparently daylighting control settings cannot be applied when using “runEnergySimulation” component.
Does anyone knows how to solve this issue?
This is something that used to encounter long ago. I even had an email exchange with David Rutten (Grasshopper creator) about it and he thought that some sort of memory limit was being exceeded after a certain number of iterations that caused it to crash. But we kinda had to chalk it up to us using Grasshopper beyond some unknown deaign limit. I usually got around the issue by breaking my parametric run up into two runs and restarting Grasshopper in between them. This always worked for me.
Can you clarify what you mean by “daylight controls”? Do you mean E+'s native daylight controls or something radiance based that you apply to the simulation through a schedule?
thanks a lot @chris for your response,
about crashing, I tried honeybee on Rhino5 on the same computer and it worked(except in 10% of the times that it still crashes). but since the total number of OSM files that I want to generate is high (about 700,000) I broke them up into 12000 and breaking it more could make the procedure a bit tedious.
about daylight controls, I get this error: