Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Installation fails - versioning of libraries needed #11

Open
gusqgm opened this issue May 2, 2023 · 4 comments
Open

Installation fails - versioning of libraries needed #11

gusqgm opened this issue May 2, 2023 · 4 comments

Comments

@gusqgm
Copy link
Collaborator

gusqgm commented May 2, 2023

Currently the main branch has no specifications for many libraries, and this is now leading to library version conflicts, as this has not been updated for a while. This issue was first reported by Jan-Daniel.

@gusqgm
Copy link
Collaborator Author

gusqgm commented May 2, 2023

As a way to overcome this issue and have a more systematic look at refactoring LSTree to be able to work with newer libraries around, I have created the refactor_2023 branch. The setup.py is already updated and should contain fixed version of all libraries that allow LSTree to run.

One last issue is still remaining which concerns tensorflow, currently installed via the installation of dl-utils. One easy way to overcome this for now is to clone the repository of dl-utils, and manually change the setup.py file there so that:

tensorflow<2.4 so install_requires list inside the setup.py from dl-utils looks like:

install_requires=[
    'numpy<=1.19.5',  # numpy 1.2 incomptabile with tensorflow 2.3, see https://stackoverflow.com/questions/66207609/notimplementederror-cannot-convert-a-symbolic-tensor-lstm-2-strided-slice0-t
    'scipy',
    'tensorflow<2.4',
    'scikit-image>=0.13',
    'scikit-learn',
    'future',
    'tqdm',
    'pytest',
    'pyyaml',
    'h5py',
    'tensorflow-addons'
]

@gusqgm
Copy link
Collaborator Author

gusqgm commented May 2, 2023

There is also an issue with pyvista. Setting pyvista==0.34.0 is not a solution because of some reshuffling in their libraries, see here for details.

One way to overcome this is to have the latest version of pyvista. I have made the corresponding changes in the setup.py from the 2023_refactor branch.

@jd-deleede
Copy link

jd-deleede commented Aug 10, 2023

Thanks Gustavo for fixing the issue above!

I found another issue while installing the refactor_2023 branch: dl-utils installs tensorflow-addons without specifying a version. This leads to the installation of tensorflow-addons version 0.19.0, which is incompatible with tensorflow 2.3. Downgrading tensorflow-addons to version 0.13.0 fixed this issue.

@gusqgm
Copy link
Collaborator Author

gusqgm commented Aug 10, 2023

Thank you @jd-deleede for spotting this and providing a solution! Currently the installations instructions are still the old ones in the main, but I will try to merge the 2023 refactor to main soon so to have package versions updated there as well.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants