Repository navigation
What's new in v1.7
BehaviorDEPOT v1.7 improves the analysis workflow without changing the standard input and output structures.
For the complete release record, see the v1.7 changelog.
The tracked-part registration dialog now shows an example tracking preview beside the mapping controls. Match each detected label to a known body part; each named part can be used once, while Other can be used for extra or unrecognized points.
Confirmed mappings are saved locally and offered as defaults when the same detected labels are encountered later. Review the mapping before confirming, especially after changing a tracking model.
Custom ROI drawing now keeps the video frame, ROI naming, redraw, and delete controls in one resizable window. In batch analyses, choose whether to reuse ROI names and limits, names only, or neither for later sessions.
- Registration and ROI windows open centered on the active display and scale to the available screen.
- A single-session result folder is named
<video-stem>_analyzed; for example,mouse01.mp4producesmouse01_analyzed. - Batch input validation reports missing or ambiguous video/tracking files before analysis begins.
- Cue-name reuse and settings import/export behavior were corrected.
- Data Exploration no longer treats the GLM intercept as a removable metric.
The v1.7 toolbox is built with MATLAB R2026a and is compatible with MATLAB releases from R2022b through R2026a Update 3. It has been tested on Windows 11 across that range. The Parallel Computing Toolbox is optional.
Legacy compatibility utilities are retained under code/deprecated. They are
not used by the current app or analysis pipeline and should not be used in new
workflows.