# Population size when creating a Study

**URL:** <https://forum.dynamobim.com/t/population-size-when-creating-a-study/96671>\
**Category:** Generative Design\
**Tags:** revit\
**Created:** [December 26, 2023, 1:56am UTC](https://forum.dynamobim.com/t/population-size-when-creating-a-study/96671 "2023-12-26T01:56:27Z")\
**Posts on this page:** 4\
**Page:** 1

<div class="post-metadata">

**Author:** ![s.iwasaki](https://avatars.discourse-cdn.com/v4/letter/s/858c86/32.png) [@s.iwasaki](https://forum.dynamobim.com/u/s.iwasaki)\
**Post date:** [December 26, 2023, 1:56am UTC](https://forum.dynamobim.com/t/population-size-when-creating-a-study/96671/1 "2023-12-26T01:56:27Z")

</div>

I have a question about setting the Population size value when creating a Study.  
The [Autodesk help page](https://help.autodesk.com/view/RVT/2023/ENU/?guid=GUID-710D71BE-3376-4647-822A-DA4D793DCBFA#GUID-710D71BE-3376-4647-822A-DA4D793DCBFA__SECTION_B1BA745FA49947E7840C682365F1B7B4) says “A rule of thumb is to set the population size to at least three times the number of inputs.”  
However, when I actually try to set it, the Study screen in Generative Design says as an issue that “Population Size must be equal or greater than 4 and a multiple of 4.”

I would like to know how to determine the value of Population size.

---

<div class="post-metadata">

**Author:** ![jacob.small](https://sea2.discourse-cdn.com/flex022/user_avatar/forum.dynamobim.com/jacob.small/32/161030_2.png) [@jacob.small](https://forum.dynamobim.com/u/jacob.small)\
**Post date:** [December 26, 2023, 2:08am UTC](https://forum.dynamobim.com/t/population-size-when-creating-a-study/96671/2 "2023-12-26T02:08:31Z")

</div>

> [@s.iwasaki](#):
>
> “A rule of thumb is to set the population size to at least three times the number of inputs.”

This is a good guideline, though in some cases you may want a multiple is ten or two or some unrelated number. It’s a rule of thumb not a steadfast certainty after all.

So as the value has to be divisible by four, round up to the next multiple of four. I.E if you have 3 inputs, multiply that by 3 to get 9, the next multiple of four is 12, so use 12 to start with. That said generally I don’t recommend dropping below 20 - again these are guidelines, and you need to learn what works for the type of problems you are solving.

---

<div class="post-metadata">

**Author:** ![s.iwasaki](https://avatars.discourse-cdn.com/v4/letter/s/858c86/32.png) [@s.iwasaki](https://forum.dynamobim.com/u/s.iwasaki)\
**Post date:** [December 26, 2023, 2:27am UTC](https://forum.dynamobim.com/t/population-size-when-creating-a-study/96671/3 "2023-12-26T02:27:01Z")

</div>

Thanks for your reply.  
Thanks to you I understand that being a multiple of 4 should be the highest priority.  
However, I do not understand why it is a multiple of 4.

Is it because when using NSGA-II, you need two families to breed?

---

<div class="post-metadata">

**Author:** ![jacob.small](https://sea2.discourse-cdn.com/flex022/user_avatar/forum.dynamobim.com/jacob.small/32/161030_2.png) [@jacob.small](https://forum.dynamobim.com/u/jacob.small)\
**Post date:** [December 26, 2023, 10:04am UTC](https://forum.dynamobim.com/t/population-size-when-creating-a-study/96671/4 "2023-12-26T10:04:15Z")

</div>

Yes. Four is a requirement of the implementation of the nsga2 algorithm.