# Shape Model Evaluation

**URL:** <https://shapeworks.discourse.group/t/shape-model-evaluation/140>\
**Category:** Uncategorized\
**Created:** [April 27, 2023, 1:04pm UTC](https://shapeworks.discourse.group/t/shape-model-evaluation/140 "2023-04-27T13:04:50Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![BMoscoloni](https://avatars.discourse-cdn.com/v4/letter/b/ed8c4c/32.png) [@BMoscoloni](https://shapeworks.discourse.group/u/BMoscoloni)\
**Post date:** [April 27, 2023, 1:04pm UTC](https://shapeworks.discourse.group/t/shape-model-evaluation/140/1 "2023-04-27T13:04:50Z")

</div>

Hi everyone,

I am working on a dataset of around 80 biventricular meshes (multi-domain), and I built a script to explore the parameter space on a subset to determine the best parameters. Until now I then opened the single scripts in Studio to look at compactness, generalization and specificity. I have tried using the shapeEvaluation modules to do this directly from script, but I am getting an error when loading the particleSystem because the files contain different number of particles (I have 5 domains). How can I go around this? Also, if there any example in the documentation on a good strategy for parameter search?

Thanks in advance

Beatrice

---

<div class="post-metadata">

**Author:** ![amorris](https://avatars.discourse-cdn.com/v4/letter/a/73ab20/32.png) [@amorris](https://shapeworks.discourse.group/u/amorris)\
**Post date:** [April 27, 2023, 10:28pm UTC](https://shapeworks.discourse.group/t/shape-model-evaluation/140/2 "2023-04-27T22:28:42Z")

</div>

Hi,

We don’t as far as I know have a good document on parameter search.

The shape evaluation and particle system classes are built around single domain processing. You will either concatenate the particle files or use a different approach. In the dev build or 6.4 release candidates, you can run `shapeworks analyze` and it will output a json file with all of these measures, which will account for multiple domains.

We still need to further develop our analysis python API and is planned.
