when installing requirements of ultrack-flow-field,I encountered Dependency conflict: geff version requirements between ultrack and traccuracy

[Reasummit]

I followed your instructions to run such commands:

conda env create -f environment_gpu.yml
conda activate ultrack-flow-field
pip install git+https://github.com/royerlab/ultrack

then I encountered Incompatibility Bug:
traccuracy 0.4.0 requires geff>=0.5.0, but you have geff 0.2.1 which is incompatible.

It seems that traccuracy 0.4.0 requires geff>=0.5.0 but ultrack 0.7.0rc1 requires geff<0.3,>=0.2.1.

Thank you for your time and for maintaining this valuable package! I appreciate any guidance you can provide.

[JoOkuma]

Hi @Reasummit , can you try installing both at the same time with

pip install git+https://github.com/royerlab/ultrack traccuracy

in your conda environment?

[Reasummit]

@JoOkuma
Sorry,but it doesnot work.After I followed your guides to run:

conda env create -f environment_gpu.yml
conda activate ultrack-flow-field
pip install git+https://github.com/royerlab/ultrack
pip install git+https://github.com/royerlab/ultrack traccuracy

Then I encountered another error when I tried to import torch:

Error loading "D:\Anaconda\envs\ultrack-flow-field\Lib\site-packages\torch\lib\fbgemm.dll" or one of its dependencies.

what should I do?

[JoOkuma]

Hi @Reasummit , sorry for the late reply, this got lost somewhere in my mail box

We are addressing the geff dependency in this PR, and this will be fixed soon #240

[JoOkuma]

The current issue is regarding torch, can you try installing with the CPU environment file to discard the possibility of being a GPU error?