# Best approach to downsample composite dataset?

**URL:** https://discourse.paraview.org/t/best-approach-to-downsample-composite-dataset/1751
**Category:** ParaView Support
**Created:** [May 2, 2019, 3:44pm UTC](https://discourse.paraview.org/t/best-approach-to-downsample-composite-dataset/1751 "2019-05-02T15:44:24Z")
**Posts on this page:** 2
**Page:** 1

<div class="post-metadata">

### Author: ![terryturton](https://discourse.paraview.org/user_avatar/discourse.paraview.org/terryturton/32/210_2.png) [@terryturton](https://discourse.paraview.org/u/terryturton)
#### Post date: [May 2, 2019, 3:44pm UTC](https://discourse.paraview.org/t/best-approach-to-downsample-composite-dataset/1751/1 "2019-05-02T15:44:24Z")

</div>

Hi, we have a composite dataset (.vtu) files and would like to downsample the data. Our solution was to merge blocks and apply a glyph filter with uniform spatial distribution since Extract Subset doesn’t work on a composite dataset. The glyph idea works but I was wondering if there were any other clever solutions that people could suggest. I would especially be interested in a solution that allows non-uniform sampling so as to sample the data in interesting regions more finely than in uninteresting regions. Thanks, Terry

---

<div class="post-metadata">

### Author: ![wascott](https://discourse.paraview.org/letter_avatar_proxy/v4/letter/w/8e8cbc/32.png) [@wascott](https://discourse.paraview.org/u/wascott)
#### Post date: [May 7, 2019, 12:00am UTC](https://discourse.paraview.org/t/best-approach-to-downsample-composite-dataset/1751/2 "2019-05-07T00:00:02Z")

</div>

Hi Terry,  
I don’t think this can be done in any cleaner of a way than you used. Although not exposed, the LOD filter just deletes points in a really rough way, that you don’t want to use. Ala