Hello,
I’m trying to install SCT (the plb/4995_new_ms_model branch), but it seems to have an issue after installing dipy.
I’m running on Ubuntu 22.04, I just cloned the repo and checked out the branch, then ran ./install_sct -y
I don’t know what changed, but I believed few weeks ago, I managed to install it successfully, but not anymore.
I attached the log file.
install_sct_log.txt (68.2 KB)
Thanks.
Edit: I believe this installation is not working as expected because when trying to import numpy from the created environment, it doesn’t find it:
lchauvin@imaging-gpu-05:/scratch/02/lchauvin/workspace/projects/spinalcordtoolbox$ ./python/envs/venv_sct/bin/python -c 'import numpy as np'
Traceback (most recent call last):
File "<string>", line 1, in <module>
ModuleNotFoundError: No module named 'numpy'
Edit 2: I also tried to follow the same steps as this test: New multi-fold MS lesion model · spinalcordtoolbox/spinalcordtoolbox@debc01f · GitHub
But I’m still getting the same error.
Edit 3: My understanding is that the installation of dipy is trying to run this code
../dipy/meson.build:106:17: ERROR: Command `/scratch/02/lchauvin/workspace/projects/spinalcordtoolbox/python/envs/venv_sct/bin/python3.10 -c '
import numpy as np
try:
incdir = os.path.relpath(np.get_include())
except Exception:
incdir = np.get_include()
print(incdir)
'` failed with status -6.
However, numpy has not been installed yet.
If I look at the tests, we can see:
Collecting dipy==1.8.0 (from -r requirements.txt (line 12))
Downloading dipy-1.8.0-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (8.6 kB)
Collecting ivadomed (from -r requirements.txt (line 16))
Downloading ivadomed-2.9.10-py3-none-any.whl.metadata (6.4 kB)
Collecting matplotlib (from -r requirements.txt (line 17))
Downloading matplotlib-3.10.6-cp310-cp310-manylinux2014_x86_64.manylinux_2_17_x86_64.whl.metadata (11 kB)
Collecting matplotlib-inline (from -r requirements.txt (line 22))
Downloading matplotlib_inline-0.1.7-py3-none-any.whl.metadata (3.9 kB)
Collecting nibabel (from -r requirements.txt (line 30))
Downloading nibabel-5.3.2-py3-none-any.whl.metadata (9.1 kB)
Collecting nilearn (from -r requirements.txt (line 31))
Downloading nilearn-0.12.1-py3-none-any.whl.metadata (9.9 kB)
Collecting nnunetv2==2.6.2 (from -r requirements.txt (line 43))
Downloading nnunetv2-2.6.2.tar.gz (211 kB)
Installing build dependencies: started
Installing build dependencies: finished with status 'done'
Getting requirements to build wheel: started
Getting requirements to build wheel: finished with status 'done'
Preparing metadata (pyproject.toml): started
Preparing metadata (pyproject.toml): finished with status 'done'
Collecting numpy<2 (from -r requirements.txt (line 45))
Downloading numpy-1.26.4-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (61 kB)
Numpy is installed after dipy.
It seems the problem is coming from GraalPy somehow that uses a different version of pip:
<frozen graalpy.pip_hook>:48: RuntimeWarning: You are using an untested version of pip. GraalPy provides patches and workarounds for a number of packages when used with compatible pip versions. We recommend to stick with the pip version that ships with this version of GraalPy.
WARNING: Disabling truststore because platform isn't supported
But I don’t understand why.
Edit 4: A collegue of mine find out that installing a new conda is messing up with the existing path. So we modified the install_sct script to use our already existing micromamba installation, and it seems to work that way. But it’s a workaround. There should be a better way to fix this.