Main repository for Padilla-Coreano Lab's projects that use Animal Pose Tracking programs such as SLEAP.
Directory Overview:
- feature_extraction_sleap.ipynb (author MC):
- calculates distances, velocities, angles etc. by creating a class for h5 files for SLEAP tracked videos (a few 2d only functions for tube test)
- feature visualization functions that displays pose position and feature value overlaid on randomly sampled frames for calculated features
- creates gifs from videos for clusters from unsupervized clustering techniques
- plots ethograms and other behavior bouts (based on cluster) graphs
- results > sleap_analysis:
- tubetest_analysis.ipynb: old notebook that plays with sleap h5 files and visualization of tracks as well as attempting to manually define a push, author MIC
- UMAP_tubetest.ipynb: attempts to do UMAP unsupervized clustering on tube test data (unsuccessfully) and attempts to make videos of clustered data
- aidan_fxn.ipynb: author AH, function to create a 2D array of start and stop frames for which a given feature passes a given threshold. The output is an array of two element arrays, where the first value is the start frame and the second value is the stop frame
- src > .gitkeep
- bin > sleap command to install sleap
- training_profiles: sleap models
- sleap_env: idk why this is here