EthoML / VAME

Variational Animal Motion Embedding - A tool for time series embedding and clustering
https://ethoml.github.io/VAME/
GNU General Public License v3.0
4 stars 0 forks source link

dev #3

Closed luiztauffer closed 1 month ago

luiztauffer commented 2 months ago
bendichter commented 2 months ago

It looks like vame/model/rnn_model.py was deleted entirely. What's the story there?

luiztauffer commented 2 months ago

looks like a git bug, but the file is there, was just moved (together with everything else) to /src image

codecov[bot] commented 2 months ago

Codecov Report

Attention: Patch coverage is 63.24201% with 161 lines in your changes are missing coverage. Please review.

Files Coverage Δ
src/vame/__init__.py 100.00% <ø> (ø)
src/vame/analysis/__init__.py 100.00% <ø> (ø)
src/vame/initialize_project/__init__.py 100.00% <ø> (ø)
src/vame/model/__init__.py 100.00% <ø> (ø)
src/vame/model/dataloader.py 96.87% <100.00%> (ø)
src/vame/model/rnn_vae.py 79.18% <100.00%> (ø)
src/vame/util/__init__.py 100.00% <ø> (ø)
src/vame/util/csv_to_npy.py 97.95% <100.00%> (ø)
src/vame/util/gif_pose_helper.py 9.72% <100.00%> (ø)
src/vame/analysis/videowriter.py 58.16% <85.71%> (ø)
... and 13 more
bendichter commented 2 months ago

Is the strategy here to create feature PRs into dev and then get client's approval before merging dev into main? That seems like a fine workflow to me, I'm just making sure we are on the same page about strategy here.

bendichter commented 2 months ago

I'd like to move from a git clone installation workflow to a pip install workflow (which I see you have already begun). Once that happens, we should not expect the user to have the demo files in an accessible location. Our tutorials should expect that the user uses vame as a pip installed library and that all of the data files are downloaded separately and placed in directories outside that library.

vinicvaz commented 2 months ago

Is the strategy here to create feature PRs into dev and then get client's approval before merging dev into main? That seems like a fine workflow to me, I'm just making sure we are on the same page about strategy here.

Yes I think it is a good strategy.

vinicvaz commented 2 months ago

I'd like to move from a git clone installation workflow to a pip install workflow (which I see you have already begun). Once that happens, we should not expect the user to have the demo files in an accessible location. Our tutorials should expect that the user uses vame as a pip installed library and that all of the data files are downloaded separately and placed in directories outside that library.

Sure, the package will not contain other folders rather than src/

bendichter commented 2 months ago

It looks like we have entirely dropped support for GPU. That's OK for now so that we can get this working cross-platform, but let's look at adding that back in the near future.