python3.sif: containerize tools (#303)

* Containerize tools
Fixes #302

* make this version 1.9.1

* bump git version

* version pins

* fix plink2 test

Author: espenhgn

CHANGELOG.md

@@ -37,6 +37,17 @@ Removed features goes here
 
 Miscellaneous goes here
 
+## [1.9.1] - 2025-03-27
+
+### Added
+
+* add fastparquet, pyarrow, dask (explicitly) packages and plink2 binary to python3.sif container
+
+### Updated 
+
+* updated scikit-learn, pandas, sksurv, seaborn packages in python3.sif container
+* updated pull_all_containers.py script
+
 ## [1.9.0] - 2025-02-20
 
 ### Added

docker/dockerfiles/python3/Dockerfile

@@ -20,6 +20,8 @@ RUN bash install_mambaforge.sh && \
 # https://www.online-utility.org/text/sort.jsp
 RUN mamba install python=3.10.6 \
     configparser=7.0.0 \
+    dask=2023.3.0 \
+    fastparquet=2024.11.0 \
     graphviz=12.2.1 \
     "h5py=3.7.0=nompi*" \
     jupyterlab=3.4.8 \
@@ -39,6 +41,7 @@ RUN mamba install python=3.10.6 \
     pandas=2.2.3 \
     pandas-plink=2.2.9 \
     psutil=5.9.3 \
+    pyarrow=19.0.1 \
     pydot=3.0.4 \
     pyreadstat=1.2.7 \
     pytables=3.7.0 \
@@ -76,6 +79,12 @@ COPY /scripts/install_plink.sh /tmp/plink/
 RUN bash /tmp/plink/install_plink.sh && \
     rm -rf /tmp/plink
 
+# Plink2
+WORKDIR /tmp/plink2
+COPY /scripts/install_plink2.sh /tmp/plink2/
+RUN bash /tmp/plink2/install_plink2.sh && \
+    rm -rf /tmp/plink2
+
 # python_convert script(s)
 WORKDIR /tools/python_convert
 RUN git clone https://github.com/precimed/python_convert.git . && \

docker/scripts/apt_get_essential.sh

@@ -16,7 +16,7 @@ apt-get update && apt-get install -y --no-install-recommends \
    dos2unix=7.4.0-2 \
    gdb=9.1-0ubuntu1 \
    gfortran=4:9.3.0-1ubuntu2 \
-   git=1:2.25.1-1ubuntu3.13 \
+   git=1:2.25.1-1ubuntu3.14 \
    less=551-1ubuntu0.3 \
    libatlas-base-dev=3.10.3-8ubuntu7 \
    libcurl4-openssl-dev=7.68.0-1ubuntu2.25 \

tests/test_python3.py

@@ -24,6 +24,7 @@
     PYTHON = f'{PREFIX} python'
     PYTHON_MOUNT = f'{PREFIX_MOUNT} python'
     PLINK = f'{PREFIX} plink'
+    PLINK2 = f'{PREFIX} plink2'
     PRSICE = f'{PREFIX} PRSice_linux'
     MINIWDL = f'{PREFIX} miniwdl'
 except FileNotFoundError:
@@ -39,6 +40,7 @@
         PYTHON = f'{PREFIX} python'
         PYTHON_MOUNT = f'{PREFIX_MOUNT} python'
         PLINK = f'{PREFIX} plink'
+        PLINK2 = f'{PREFIX} plink2'
         PRSICE = f'{PREFIX} PRSice_linux'
         MINIWDL = f'{PREFIX} miniwdl'
     except FileNotFoundError:
@@ -47,6 +49,7 @@
         PYTHON = 'python'
         PYTHON_MOUNT = 'python'
         PLINK = 'plink'
+        PLINK2 = 'plink2'
         PRSICE = 'PRSice_linux'
         MINIWDL = 'miniwdl'
 
@@ -55,6 +58,12 @@ def test_python3_plink():
     out = subprocess.run(call.split(' '), check=False)
     assert out.returncode == 0
 
+def test_python3_plink2():
+    """test plink2"""
+    call = f'{PLINK2} --version'
+    out = subprocess.run(call.split(' '), check=False)
+    assert out.returncode == 0
+
 def test_python3_prsice():
     call = f'{PRSICE} --version'
     out = subprocess.run(call.split(' '), check=False)
@@ -95,7 +104,8 @@ def test_python3_miniwdl():
 def test_python3_packages():
     packages = [
         'configparser',
-        'pydot',
+        'dask',
+        'fastparquet',
         'dxpy',
         'graphviz',
         'h5py',
@@ -113,6 +123,8 @@ def test_python3_packages():
         'pandas',
         'pandas_plink',
         'plinkio',
+        'pyarrow',
+        'pydot',
         'pyliftover',
         'pyreadstat',
         'redcap',  # pycap

version/version.py

@@ -5,7 +5,7 @@
 _PATCH = "1"
 # This is mainly for nightly builds which have the suffix ".dev$DATE". See
 # https://semver.org/#is-v123-a-semantic-version for the semantics.
-_SUFFIX = "dev"
+_SUFFIX = ""
 
 VERSION_SHORT = "{0}.{1}".format(_MAJOR, _MINOR)
 VERSION = "{0}.{1}.{2}{3}".format(_MAJOR, _MINOR, _PATCH, _SUFFIX)