[ENH] update docker image

[Remi-Gau]

• change of base distro
• update python to 3.10
• update bids validator to 1.9.9
• update and pin version packages for defacing

[Remi-Gau]

closes #68

[Remi-Gau]

blocked by ReproNim/neurodocker#532

[Remi-Gau]

with this

docker run --rm repronim/neurodocker:0.9.5 generate docker \
             --base-image bids/base_validator:1.11.0 \
             --yes \
             --pkg-manager apt \
             --install git num-utils gcc g++ curl build-essential nano\
             --miniconda \
                version=latest \
                conda_install="python=3.10 numpy nipype nibabel pandas datalad git-annex" \
                pip_install="deepdefacer scikit-learn tensorflow scikit-image pydeface==2.0.2 nobrainer==0.4.0 quickshear==1.2.0 datalad-osf" \
             --run-bash "mkdir -p /opt/nobrainer/models && cd /opt/nobrainer/models && datalad clone https://github.com/neuronets/trained-models && cd trained-models && git-annex enableremote osf-storage && datalad get -s osf-storage ." \
             --run-bash "git clone https://github.com/mih/mridefacer" \
             --env MRIDEFACER_DATA_DIR=/mridefacer/data \
             --run-bash "mkdir /home/mri-deface-detector && cd /home/mri-deface-detector && npm install sharp --unsafe-perm && npm install -g mri-deface-detector --unsafe-perm && cd ~" \
             --run-bash "git clone https://github.com/miykael/gif_your_nifti && cd gif_your_nifti && source activate bidsonym && python setup.py install" \
             --fsl version=6.0.6.4 method=binaries \
             --copy . /home/bm \
             --run-bash "chmod a+x /home/bm/bidsonym/fs_data/mri_deface" \
             --run-bash "source activate bidsonym && cd /home/bm && pip install -e ." \
             --env IS_DOCKER=1 \
             --workdir '/tmp/' \
             --entrypoint "/neurodocker/startup.sh  bidsonym"

we get:

#0 2681.5 Collecting sklearn (from deepdefacer)
#0 2681.5   Downloading sklearn-0.0.post5.tar.gz (3.7 kB)
#0 2681.5   Preparing metadata (setup.py): started
#0 2681.7   Preparing metadata (setup.py): finished with status 'error'
#0 2681.7   error: subprocess-exited-with-error
#0 2681.7   
#0 2681.7   × python setup.py egg_info did not run successfully.
#0 2681.7   │ exit code: 1
#0 2681.7   ╰─> [18 lines of output]
#0 2681.7       The 'sklearn' PyPI package is deprecated, use 'scikit-learn'
#0 2681.7       rather than 'sklearn' for pip commands.
#0 2681.7       
#0 2681.7       Here is how to fix this error in the main use cases:
#0 2681.7       - use 'pip install scikit-learn' rather than 'pip install sklearn'
#0 2681.7       - replace 'sklearn' by 'scikit-learn' in your pip requirements files
#0 2681.7         (requirements.txt, setup.py, setup.cfg, Pipfile, etc ...)
#0 2681.7       - if the 'sklearn' package is used by one of your dependencies,
#0 2681.7         it would be great if you take some time to track which package uses
#0 2681.7         'sklearn' instead of 'scikit-learn' and report it to their issue tracker
#0 2681.7       - as a last resort, set the environment variable
#0 2681.7         SKLEARN_ALLOW_DEPRECATED_SKLEARN_PACKAGE_INSTALL=True to avoid this error
#0 2681.7       
#0 2681.7       More information is available at
#0 2681.7       https://github.com/scikit-learn/sklearn-pypi-package
#0 2681.7       
#0 2681.7       If the previous advice does not cover your use case, feel free to report it at
#0 2681.7       https://github.com/scikit-learn/sklearn-pypi-package/issues/new
#0 2681.7       [end of output]
#0 2681.7   
#0 2681.7   note: This error originates from a subprocess, and is likely not a problem with pip.
#0 2681.7 error: metadata-generation-failed
#0 2681.7 
#0 2681.7 × Encountered error while generating package metadata.
#0 2681.7 ╰─> See above for output.
#0 2681.7 
#0 2681.7 note: This is an issue with the package mentioned above, not pip.
#0 2681.7 hint: See above for details.
------
ERROR: failed to solve: executor failed running [/bin/sh -c apt-get update -qq     && apt-get install -y -q --no-install-recommends            bzip2            ca-certificates            curl     && rm -rf /var/lib/apt/lists/*     && export PATH="/opt/miniconda-latest/bin:$PATH"     && echo "Downloading Miniconda installer ..."     && conda_installer="/tmp/miniconda.sh"     && curl -fsSL -o "$conda_installer" https://repo.continuum.io/miniconda/Miniconda3-latest-Linux-x86_64.sh     && bash "$conda_installer" -b -p /opt/miniconda-latest     && rm -f "$conda_installer"     && conda update -yq -nbase conda     && conda config --system --prepend channels conda-forge     && conda config --set channel_priority strict     && conda config --system --set auto_update_conda false     && conda config --system --set show_channel_urls true     && conda init bash     && conda install -y  --name base            "python=3.10"            "numpy"            "nipype"            "nibabel"            "pandas"            "datalad"            "git-annex"     && bash -c "source activate base     &&   python -m pip install --no-cache-dir               "deepdefacer"              "scikit-learn"              "tensorflow"              "scikit-image"              "pydeface==2.0.2"              "nobrainer==0.4.0"              "quickshear==1.2.0"              "datalad-osf""     && sync && conda clean --all --yes && sync     && rm -rf ~/.cache/pip/*]: exit code: 1

seems deepfacer is using an antiquated version of scikit-learn

[yarikoptic]

Hey @Remi-Gau -- did you approach this since then or what do you use for your defacing needs these days? ;-)

[Remi-Gau]

nope: not retried

[PeerHerholz]

Hi @Remi-Gau,

are you in CET or EST right now? We could chat one of these, as I would also like to solve and ideally move this to workflows.
Otherwise during the brainhack (virtually or in-person)?

Thanks!

Cheers, Peer

[Remi-Gau]

Brussel based. Could chat during the OHBM period.