# Cell Grid
**URL:** https://discourse.paraview.org/t/cell-grid/17178
**Category:** Development
**Created:** [September 10, 2025, 2:38pm UTC](https://discourse.paraview.org/t/cell-grid/17178 "2025-09-10T14:38:47Z")
**Posts on this page:** 4
**Page:** 1
### Author:  [@Frankx9](https://discourse.paraview.org/u/Frankx9)
#### Post date: [September 10, 2025, 2:38pm UTC](https://discourse.paraview.org/t/cell-grid/17178/1 "2025-09-10T14:38:47Z")
Hi,
could you point me out to where do I find a file example of the novel CellGrid that is in Paraview that I can potentially download and inspect?
Any further documentation that I might look at while trying to implement it?
Best
---
### Author:  [@Lucas\_Givord](https://discourse.paraview.org/u/Lucas_Givord)
#### Post date: [September 11, 2025, 7:16am UTC](https://discourse.paraview.org/t/cell-grid/17178/2 "2025-09-11T07:16:31Z")
I believe @dcthomp could help you
---
### Author:  [@dcthomp](https://discourse.paraview.org/u/dcthomp)
#### Post date: [September 11, 2025, 1:35pm UTC](https://discourse.paraview.org/t/cell-grid/17178/3 "2025-09-11T13:35:27Z")
Sorry for the slow reply. Here are some examples
- [dgHexahedra.dg](https://discourse.paraview.org/uploads/short-url/9IrlQ9sFXgd3FRnBT6zhG5Mcsef.dg) (5.7 KB) has two tri-linear (in shape) hexahedra with functions from a variety of different function spaces (including tri-quadratic Lagrange/H-grad). The `scalar3` field shows how continuous (point-centered H-grad a.k.a. Lagrange) fields can be specified. Other fields (`scalar0`, `scalar1`, `scalar2`, `quadratic`) show how discontinuous H-grad fields may be specified. The `curl1` and `div1` fields show how the fields from the H-curl and H-div function spaces may be specified (these are only for discontinuous formulations at the moment).
- [dgGhostHexahedra.dg](https://discourse.paraview.org/uploads/short-url/sSCBSC9Pa4MSxsYd9cDuF51r9FL.dg) (5.6 KB) has the same geometry above but demonstrates how `ghost-node` information in the `cell-types` dictionary entries can be used to handle distributed-memory parallelism. 8 of the 12 nodes are marked as ghost nodes, so only 5 faces of one cell appear when rendering the data.
- [dgTetrahedra.dg](https://discourse.paraview.org/uploads/short-url/waDnuOKdeUeIBgNhexzB7rAXQTk.dg) (5.1 KB) has two linear tetrahedra.
- [dgQuadrilateral.dg](https://discourse.paraview.org/uploads/short-url/8vBv4XUlobygPCdmJGKStHjpyoK.dg) (3.8 KB) has two bi-linear quadrilaterals.
- [dgQuadraticQuadrilaterals.dg](https://discourse.paraview.org/uploads/short-url/pS9kQrVsfOHhrxgIr1y5YDHU6gv.dg) (4.7 KB) has two bi-quadratic (in shape) quadrilaterals. You will notice some futziness with the adaptive (on-GPU) tessellation as you align the camera with some axes. But the rendering does perform adaptive tessellation and per-fragment scalar-field interpolation for coloring.
- [dgTriangle.dg](https://discourse.paraview.org/uploads/short-url/wivovWN0CXK2QEZLCusX1k3BW9S.dg) (3.7 KB)
- [dgPyr19.dg](https://discourse.paraview.org/uploads/short-url/6paQttIkVETiSqcAHvG5HNzkcVn.dg) (2.5 KB) demonstrates a 19-node pyramid.
- [dgWedges.dg](https://discourse.paraview.org/uploads/short-url/kQS4XcooBNvDUHhGuAMFxBfVh2k.dg) (3.9 KB) has two tri-linear wedge elements.
- [dgMixed.dg](https://discourse.paraview.org/uploads/short-url/ct96y0WqBVjy60YsbQJlsTPNVkK.dg) (6.0 KB) demonstrates how multiple types of cells may exist in the same vtkCellGrid instance (though I strongly advise against this: it is preferred to use a `vtkPartitionedDataSetCollection` of `vtkCellGrid` instances with one cell-grid per cell shape).
- [fandisk.dg](https://discourse.paraview.org/uploads/short-url/ck9ek14ytjrohCzFArBteV1dAts.dg) (110.0 KB) is a larger, all-hexahedral mesh with a few nonsensical scalar fields.
- [bunny-small.dg](https://discourse.paraview.org/uploads/short-url/rkdRTOk18GtVJ0VZWzjzHtpCS4l.dg) (7.0 MB) is a larger (though still not large), all-tetrahedral volume mesh of the Stanford bunny.
In terms of developer documentation:
- if you are trying to understand the JSON cell-grid file format [see here](https://docs.vtk.org/en/latest/modules/vtk-modules/IO/CellGrid/README.html)
- if you are writing filters for cell-grids [see here](https://docs.vtk.org/en/latest/modules/vtk-modules/Filters/CellGrid/README.html)
- if you are writing (or trying to read the existing) rendering code for cell-grids [see here](https://docs.vtk.org/en/latest/modules/vtk-modules/Rendering/CellGrid/README.html)
- discourse topics provide a brief overview of the design [here](https://discourse.vtk.org/t/discontinuous-galerkin-elements-and-other-novel-cell-types-function-spaces/9209) and [here](https://discourse.vtk.org/t/new-vtkcellgrid-functionality/14640)
- There are a couple of figures in [this blog post](https://www.kitware.com/novel-discretization-and-formulation-support-in-vtk-and-paraview/) that might be helpful.
That’s about it. Feel free to ask more specific, detail-oriented questions.
---
### Author:  [@Frankx9](https://discourse.paraview.org/u/Frankx9)
#### Post date: [September 12, 2025, 8:50am UTC](https://discourse.paraview.org/t/cell-grid/17178/4 "2025-09-12T08:50:48Z")
Hi @dcthomp thanks a lot for the very detailed answer. I’ll try to mess around with it a little bit and if any doubt arises I will get back in touch!