# Filter to detect and treat outliers in field data

**URL:** https://discourse.paraview.org/t/filter-to-detect-and-treat-outliers-in-field-data/16272
**Category:** ParaView Support
**Created:** [March 7, 2025, 5:46pm UTC](https://discourse.paraview.org/t/filter-to-detect-and-treat-outliers-in-field-data/16272 "2025-03-07T17:46:05Z")
**Posts on this page:** 3
**Page:** 1

<div class="post-metadata">

### Author: ![simonw](https://discourse.paraview.org/letter_avatar_proxy/v4/letter/s/34f0e0/32.png) [@simonw](https://discourse.paraview.org/u/simonw)
#### Post date: [March 7, 2025, 5:46pm UTC](https://discourse.paraview.org/t/filter-to-detect-and-treat-outliers-in-field-data/16272/1 "2025-03-07T17:46:05Z")

</div>

I have 2d measurement data (x,y,u) which have some outliers, see e.g the highlighted regions below. To reproduce my picture, Table2Points and Delaunay2d needs to be applied to the attached file.

Is there any filter to detect and ideally replaces the outliers with better values?

 ![pic](https://discourse.paraview.org/uploads/default/original/3X/4/e/4ebd62c2c23ab4e58d1e852b30b843a8963ddef3.png)

[data.txt](https://discourse.paraview.org/uploads/short-url/bAViyaMB2bnWZPpR9ZhGMJZKCmO.txt) (611.0 KB)

---

<div class="post-metadata">

### Author: ![mwestphal](https://discourse.paraview.org/user_avatar/discourse.paraview.org/mwestphal/32/17_2.png) [@mwestphal](https://discourse.paraview.org/u/mwestphal)
#### Post date: [March 9, 2025, 11:09am UTC](https://discourse.paraview.org/t/filter-to-detect-and-treat-outliers-in-field-data/16272/2 "2025-03-09T11:09:54Z")

</div>

The better approach is probably a simple programmable filter

---

<div class="post-metadata">

### Author: ![simonw](https://discourse.paraview.org/letter_avatar_proxy/v4/letter/s/34f0e0/32.png) [@simonw](https://discourse.paraview.org/u/simonw)
#### Post date: [March 9, 2025, 3:35pm UTC](https://discourse.paraview.org/t/filter-to-detect-and-treat-outliers-in-field-data/16272/3 "2025-03-09T15:35:38Z")

</div>

Any recommendation what command woule be pertinent here?
