Hello tobac users,
We are happy to announce the release of v1.3.3 today. This release is a hotfix designed to fix a bug that appears when performing predictive tracking (i.e., `method_linking=’predict’ in tobac.linking_trackpy), and we expect that this fix will substantially improve the tracking performance when using this option.
These changes are workarounds within tobac to address issues upstream in Trackpy predictive tracking. Trackpy (
http://soft-matter.github.io/trackpy/v0.5.0/) is a library that tobac uses to perform tracking on the detected features. We’ve unfortunately discovered an issue with predictive tracking, which was introduced in 2018. If you are interested in learning more, we encourage you to reach out or read the discussions on GitHub here:
https://github.com/soft-matter/trackpy/issues/699 and here:
https://github.com/tobac-project/tobac/issues/169. Based on our experiences, the poor performance of predictive tracking will most impact datasets with large cell displacements due to fast movement within a timestep.
We have begun the process of addressing these changes upstream in Trackpy, but we have identified a workaround for these issues. The changes we are introducing in v1.3.3 allow the predictive tracking method to perform as originally intended.
We
strongly recommend updating tobac immediately if you use predictive tracking, and the release notes are included below. Note that your tracking output will likely change from v1.3.2 and before to v1.3.3. If maintaining consistency in your tracking output is important to you, we recommend that you stay on v1.3.2 for now.
Special thanks to Nils Pfeifer at TROPOS, Leipzig and Kelcy Brunner at Texas Tech for their quick work to address this issue.
Release notes:
Changelog:
Bug fixes- Added a workaround to a bug in trackpy that fixes predictive tracking #170
Thanks,
The tobac dev team