# Python Seg Fault
**URL:** https://discourse.vtk.org/t/python-seg-fault/3353
**Category:** Development
**Created:** [May 20, 2020, 4:46pm UTC](https://discourse.vtk.org/t/python-seg-fault/3353 "2020-05-20T16:46:31Z")
**Posts on this page:** 4
**Page:** 1
### Author:  [@gatti](https://discourse.vtk.org/u/gatti)
#### Post date: [May 20, 2020, 4:46pm UTC](https://discourse.vtk.org/t/python-seg-fault/3353/1 "2020-05-20T16:46:31Z")
Hi,
I’m messaging because I believe I came across a bug. I’ve found a hacky fix for it, but I don’t think my fix is the correct long-term solution. Im using vtk==8.2.0 installed from Conda.
If you `filter.GetOutput()` from some filter and then try to delete a cell `polydata.DeleteCell(idx)` of the outputted polydata you get a seg fault. However, if you simply interact with the cells in some (any?) way and then perform the delete there is no error (e.g., by doing `polydata.GetCell(idx)` and then do `polydata.DeleteCell(idx)`; note that I don’t think idx has to be the same between these, and `.GetCell()` only needs to be done once and then you can iterate over the entire mesh and delete whatever cells you want.
Basic/reproducible example code:
ERROR:
```auto
import vtk
sphereSource = vtk.vtkSphereSource()
sphereSource.SetCenter(0.0, 0.0, 0.0)
sphereSource.SetRadius(5.0)
sphereSource.SetPhiResolution(100)
sphereSource.SetThetaResolution(100)
sphereSource.Update()
sphere = vtk.vtkPolyData()
sphere.DeepCopy(sphereSource.GetOutput())
sphere.DeleteCell(1)
```
RESOLUTION:
```auto
import vtk
sphereSource = vtk.vtkSphereSource()
sphereSource.SetCenter(0.0, 0.0, 0.0)
sphereSource.SetRadius(5.0)
sphereSource.SetPhiResolution(100)
sphereSource.SetThetaResolution(100)
sphereSource.Update()
sphere = vtk.vtkPolyData()
sphere.DeepCopy(sphereSource.GetOutput())
sphere.GetCell(1)
sphere.DeleteCell(1)
```
Note that it doesn’t matter if I `DeepCopy` the output of the filter or not.
Clearly, the fix is extremely simple and I can easily include this in any of my code. However, without knowing this is the problem it could be a pain to sort out.
Let me know if any additional information is of use.
Thanks,
Anthony.
---
### Author:  [@dgobbi](https://discourse.vtk.org/u/dgobbi)
#### Post date: [May 20, 2020, 4:59pm UTC](https://discourse.vtk.org/t/python-seg-fault/3353/2 "2020-05-20T16:59:42Z")
Certain methods require that `BuildLinks()` is called beforehand. See the [vtkPolyData](https://vtk.org/doc/nightly/html/classvtkPolyData.html) docs for more info.
```auto
sphere.EditableOn()
sphere.BuildLinks()
sphere.DeleteCell(1)
```
Edit: The polydata should be marked Editable, too, so I’ve added that.
---
### Author:  [@gatti](https://discourse.vtk.org/u/gatti)
#### Post date: [May 20, 2020, 5:09pm UTC](https://discourse.vtk.org/t/python-seg-fault/3353/3 "2020-05-20T17:09:16Z")
Thanks, David.
Appreciate the prompt reply. Im assuming that by calling `.GetCell()` that `.BuildLinks()` is called as a part of it - hence why my hack works.
I was going to ask something about having `.BuildLinks()` run by default with `.GetOutput()`, but from reading through the docs more it seems like some functions require it isn’t called yet.
I guess something you just need to learn once along the path to know to look for it as something that is needed to be done.
---
### Author:  [@dgobbi](https://discourse.vtk.org/u/dgobbi)
#### Post date: [May 20, 2020, 5:32pm UTC](https://discourse.vtk.org/t/python-seg-fault/3353/4 "2020-05-20T17:32:48Z")
BuildLinks() isn’t called automatically because it is expensive and is only required in certain situations.