# CollocationPoints.jl --- Generating variable-density points for SciML

**URL:** <https://discourse.julialang.org/t/collocationpoints-jl-generating-variable-density-points-for-sciml/130125>\
**Category:** Package Announcements\
**Tags:** announcement\
**Created:** [June 23, 2025, 8:42am UTC](https://discourse.julialang.org/t/collocationpoints-jl-generating-variable-density-points-for-sciml/130125 "2025-06-23T08:42:16Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![pkmishra](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/pkmishra/32/14238_2.png) [@pkmishra](https://discourse.julialang.org/u/pkmishra)\
**Post date:** [June 23, 2025, 8:42am UTC](https://discourse.julialang.org/t/collocationpoints-jl-generating-variable-density-points-for-sciml/130125/1 "2025-06-23T08:42:16Z")

</div>

I have been porting my older utility codes to Julia. Here is something I used for generating nodes for solving PDEs with meshfree methods and sparse parameterisation of inverse problems. This may be useful in solving PDEs with scientific machine learning as well. It’s not a registered package yet

> **[GitHub - pankajkmishra/CollocationPoints.jl: Julia package for generating collocation points...](https://github.com/pankajkmishra/CollocationPoints.jl)**
>
> Julia package for generating collocation points inside various domains, with support for variable point density. The primary goal here is to use this for solving PDEs with physics-informed machine learning

---

<div class="post-metadata">

**Author:** ![juliohm](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/juliohm/32/215266_2.png) [@juliohm](https://discourse.julialang.org/u/juliohm)\
**Post date:** [June 23, 2025, 10:46am UTC](https://discourse.julialang.org/t/collocationpoints-jl-generating-variable-density-points-for-sciml/130125/2 "2025-06-23T10:46:09Z")

</div>

@pkmishra are you aware that Meshes.jl already supports homogeneous, density-based and clustered sampling in 2D and 3D meshes? You can see some examples here:

> **[Point processes · GeoStats.jl](https://juliaearth.github.io/GeoStatsDocs/stable/pointprocs/)**
>
> Documentation for GeoStats.jl.

More generally, these are called random point processes.