# Reuse weighting function and sensitivity matrix

**URL:** <https://simpeg.discourse.group/t/reuse-weighting-function-and-sensitivity-matrix/133>\
**Category:** Potential Fields\
**Created:** [August 20, 2020, 6:10am UTC](https://simpeg.discourse.group/t/reuse-weighting-function-and-sensitivity-matrix/133 "2020-08-20T06:10:52Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![Alexey](https://avatars.discourse-cdn.com/v4/letter/a/47e85d/32.png) [@Alexey](https://simpeg.discourse.group/u/Alexey)\
**Post date:** [August 20, 2020, 6:10am UTC](https://simpeg.discourse.group/t/reuse-weighting-function-and-sensitivity-matrix/133/1 "2020-08-20T06:10:52Z")

</div>

Is it possible to reuse calculated weights and sensitivity matrix to restart the inversion with new parameters (alphas, norms, bounds etc)?  
Now I noticed that if SimPEG finds the created zarr-file, it will be reloaded. Is this correct?

---

<div class="post-metadata">

**Author:** ![domfournier](https://avatars.discourse-cdn.com/v4/letter/d/a3d4f5/32.png) [@domfournier](https://simpeg.discourse.group/u/domfournier)\
**Post date:** [August 20, 2020, 2:30pm UTC](https://simpeg.discourse.group/t/reuse-weighting-function-and-sensitivity-matrix/133/2 "2020-08-20T14:30:28Z")

</div>

Hi @Alexey,  
Yes, you are correct. As long as you keep the data and mesh the same, then the linear forward operator will have the same shape, SimPEG will automatically re-load the zarr file and skip the sensitivity part. This is true for both potential fields (Grav/Mag), even the non-linear MVI.

If you to change the data and/or the mesh, and they happen to have the exact same dimensions (it can happen), you need to make sure to manually delete the zarr on disk.

Happy inversion.