How to create a new name for each study based on input parameters?

[francisco.contreras]

Hi Mostapha,

My questions is actually with step 1). How can I create all the design options? The problem I have is: that with the “exporttoOpenstudio” block, I need a new name each time otherwise the new simulation will replace the previous one.

Any ideas?

Thanks in advance.

[LUISARTUROREYESVALEN]

You can use colibri just to write the files by including and concatenating the variable names so each iteration is named in a different way. I hope the image explains it.

Best
Arturo

[francisco.contreras]

Thank you very much, this really worked.

[Ramana]

Hello @mostapha

I am running a parametric study with more than a 10000 iterations with 4 minutes of runtime each. In the end, I wish to minimize EUI and find the exact inputs which gives me the minimum EUI. Currently my model is too huge and I want to make use of the parallel computing power of my CPU. But the ModelToOSM component deletes the previous .osm file as it writes the new one. I cannot find the possibility to change the filename of the generated .osm file in the current component.

Can you please tell me how to setup my model correctly to achieve my goal?

Thanks in advance!

sample (2).gh (52.0 KB)

[AbrahamYezioro]

You need to set the folder name according to whatever you need to be to reflect the parameters values.
The concatenate component can be your friend here.
From your image, your using the default folder. As a result each new case is overwriting the previous one.
You can not change the file name.

-A.

[Ramana]

Thanks for your response. I used concatenate component as suggested as shown in the image below:

I also generated the folders with the .osm files.

But the RunOSM component then does not work directly. Should I then manually open each folder and collect all the .osm files to a single folder before I run the RunOSM component? Even if I do that, how would I link back my input parameters to the outputs that is currently being generated?

[AbrahamYezioro]

The concatenate component is your friend again here, like so:

I believe the image is self explaining.
The panel in blue is one of the cases saved in a folder (i.e. Case X), just to compare to the concatenate output.

-A.

[Ramana]

Hey @AbrahamYezioro

Thanks again for this idea.

I did the same for 5 iteration scenario and the parallel computing just worked fine. But when I increased the input to 95 iterations with 40 parallel cores (I selected 40 because my desktop has 40 logical processors), the simulation RunOSM component failed. The error message also just reads as: “1. Solution exception:At least one error has occurred.”

And, I am still not sure how to link the outputs generated (EUI) with the inputs (Window ratio), i.e. After parallel computing, I want an excel sheet with 95 Window to wall ratios with the corresponding EUI. How should I setup my Colibri for that?

Best
Ramana