Prepare for python 3.10

.github/workflows/build.yml

@@ -0,0 +1,66 @@
+# This workflow will try to build and install the software in different ways.
+
+name: Build and installation tests
+
+on:
+  push:
+    branches: ['master', 'devel']
+  pull_request:
+    branches: '*'
+
+jobs:
+  pure-pip-installation:
+    # This stage only tests if the installation is possible.
+    # The evironment created herein will be discared and re-created in the test stage.
+    runs-on: ubuntu-latest
+    strategy:
+      matrix:
+      # Add multiple Python versions here to run tests on new(er) versions.
+        python-version: ["3.10"]
+      fail-fast: false
+    steps:
+    - uses: actions/checkout@v3
+    - name: Set up Python ${{ matrix.python-version }}
+      uses: actions/setup-python@v3
+      with:
+        python-version: ${{ matrix.python-version }}
+    - name: Build and install
+      run: |
+        python -m pip install --upgrade pip
+        # Install with -e (in editable mode) to allow the tracking of the test coverage
+        pip install -e .
+        # Check result of installation
+        python -c "import loadskernel"
+        which loads-kernel
+        which model-viewer
+        which loads-compare
+
+  conda-and-pip-installation:
+    # This stage only tests if the installation is possible.
+    # The evironment created herein will be discared and re-created in the test stage.
+    runs-on: ubuntu-latest
+    strategy:
+      matrix:
+      # Add multiple Python versions here to run tests on new(er) versions.
+        python-version: ["3.10", "3.11"]
+      fail-fast: false
+    steps:
+    - uses: actions/checkout@v3
+    - name: Set up Python ${{ matrix.python-version }}
+      uses: actions/setup-python@v3
+      with:
+        python-version: ${{ matrix.python-version }}
+    - name: Build and install
+      run: |
+        # Install same requirements as to be used during regression testing
+        source $CONDA/etc/profile.d/conda.sh
+        conda activate
+        conda install -y -c conda-forge --file ./tests/requirements.txt
+        # Install with -e (in editable mode) to allow the tracking of the test coverage
+        pip install -e .
+        # Check result of installation
+        python -c "import loadskernel"
+        which loads-kernel
+        which model-viewer
+        which loads-compare
+

.github/workflows/coding-style.yml

@@ -17,8 +17,8 @@ jobs:
     runs-on: ubuntu-latest
     strategy:
       matrix:
-        # Add multiple Python versions here to run tests on new(er) versions.
-        python-version: ["3.8"]
+        # Add the Python versions here to run tests on new(er) versions.
+        python-version: ["3.11"]
     steps:
     - uses: actions/checkout@v3
     - name: Set up Python ${{ matrix.python-version }}
@@ -40,8 +40,8 @@ jobs:
     runs-on: ubuntu-latest
     strategy:
       matrix:
-      # Add multiple Python versions here to run tests on new(er) versions.
-        python-version: ["3.8"]
+        # Add multiple Python versions here to run tests on new(er) versions.
+        python-version: ["3.11"]
     steps:
     - uses: actions/checkout@v3
     - name: Set up Python ${{ matrix.python-version }}
@@ -50,10 +50,14 @@ jobs:
         python-version: ${{ matrix.python-version }}
     - name: Install dependencies
       run: |
-        python -m pip install --upgrade pip
-        pip install pylint
+        # Install same requirements as to be used during regression testing
+        source $CONDA/etc/profile.d/conda.sh
+        conda activate
+        conda install -y -c conda-forge --file ./tests/requirements.txt
         # Install the package itself to make sure that all imports work.
         pip install .
     - name: Analysing the code with pylint
       run: |
+        source $CONDA/etc/profile.d/conda.sh
+        conda activate
         pylint $(git ls-files '*.py') --fail-under=7.0

.github/workflows/regression-tests.yml

@@ -1,6 +1,4 @@
-# This workflow will install and then lint the code with Flake8 and Pylint.
-# For more information see: https://docs.github.com/en/actions/automating-builds-and-tests/building-and-testing-python
-
+# This workflow will run some regression tests.
 name: Regression Tests
 
 on:
@@ -10,36 +8,51 @@ on:
     branches: '*'
 
 jobs:
-  pip-installation:
-    # This stage only tests if the installation is possible.
-    # The evironment created herein will be discared and re-created in the test stage.
+
+  Pytest:
     runs-on: ubuntu-latest
     strategy:
       matrix:
-      # Add multiple Python versions here to run tests on new(er) versions.
-        python-version: ["3.8"]
+        # Add multiple Python versions here to run tests on new(er) versions.
+        python-version: ["3.10"]
     steps:
     - uses: actions/checkout@v3
     - name: Set up Python ${{ matrix.python-version }}
       uses: actions/setup-python@v3
       with:
         python-version: ${{ matrix.python-version }}
-    - name: Build and install
+    - name: Install dependencies
       run: |
         python -m pip install --upgrade pip
-        # Install with -e (in editable mode) to allow the tracking of the test coverage
+        # Install the package itself to make sure that all imports work.
         pip install -e .
-        # Check result of installation
-        which loads-kernel
-        which model-viewer
-        which loads-compare
+    - name: Analysing the code with pytest
+      run: |
+        # Run the actual testing of the code with pytest
+        # Using python -m pytest is necessary because pytest has the habit of not looking in the site-packages of the venv
+        python -m pytest -v --basetemp=./tmp -k test_dummy --cov=loadskernel  --cov=modelviewer --cov=loadscompare --junitxml=testresult.xml
+        # Create some reports
+        coverage report
+        coverage xml -o coverage.xml
+        coverage html --directory ./coverage
+    - name: Upload test restults and coverage as an artifact
+      uses: actions/upload-artifact@v4
+      with:
+        name: test results and coverage
+        path: |
+          testresult.xml
+          coverage.xml
+          coverage
+        if-no-files-found: ignore
 
   Jupyter:
+    # Building the Jupyter book is not really a regression test. However, it has to be in this workflow due to the handling of 
+    # the artifacts.
     runs-on: ubuntu-latest
     strategy:
       matrix:
-        # Add multiple Python versions here to run tests on new(er) versions.
-        python-version: ["3.8"]
+        # Select Python version to be used for compiling here.
+        python-version: ["3.10"]
     steps:
     - uses: actions/checkout@v3
     - name: Set up Python ${{ matrix.python-version }}
@@ -49,7 +62,7 @@ jobs:
     - name: Install dependencies
       run: |
         python -m pip install --upgrade pip
-        pip install jupyter-book
+        # Install the package itself to make sure that all imports work.
         pip install .
     - name: Assemble the tutorials to a jupyter book and build htlm pages
       run: |
@@ -58,23 +71,37 @@ jobs:
         mkdir ./doc/html
         mv ./doc/tutorials/_build/html ./doc/html/tutorials
     - name: Upload Jupyter book as an artifact
-      uses: actions/upload-artifact@v3
+      uses: actions/upload-artifact@v4
       with:
         name: tutorials
         path: ./doc/html
         if-no-files-found: ignore
-    - name: Upload Jupyter book for pages
+
+  combine-pages:
+    runs-on: ubuntu-latest
+    # Add a dependency to the build job
+    needs: [Jupyter, Pytest]
+    steps:
+    - uses: actions/download-artifact@v4
+      with:
+        merge-multiple: true
+    - name: See what we've got and merge artifacts
+      run: |
+        ls -la
+        mkdir pages
+        mv ./tutorials ./pages/tutorials
+        mv ./coverage ./pages/coverage
+    - name: Upload artifact for pages
       # This is not a normal artifact but one that can be deployed to the GitHub pages in the next step
       uses: actions/upload-pages-artifact@v3
       with:
-        name: github-pages # This name may not be changed according to the documentation
-        path: ./doc/html
+        name: github-pages  # This name may not be changed according to the documentation
+        path: ./pages  # There must be only one path
         if-no-files-found: ignore
-
+    
   deploy-pages:
     # Add a dependency to the build job
-    needs: Jupyter
-
+    needs: combine-pages
     # Grant GITHUB_TOKEN the permissions required to make a Pages deployment
     permissions:
       pages: write      # to deploy to Pages

.gitlab-ci.yml

@@ -7,7 +7,7 @@ stages:
   - deploy
 
 .virtenv: &virtualenv
-  - source /work/voss_ar/Software/miniconda3/etc/profile.d/conda.sh
+  - source /work/f_jwsb/software/miniconda-3.11/etc/profile.d/conda.sh
   - conda activate
   # To make things faster, re-use existing site packages. 
   - python -m venv virtualenv --system-site-packages
@@ -17,9 +17,7 @@ stages:
   # Check python version
   - which python
   - which pytest
-  # Set-up MPI
-  - export PATH=/work/voss_ar/Software/mpich-3.4.2/bin:$PATH
-  - export LD_LIBRARY_PATH=/work/voss_ar/Software/mpich-3.4.2/lib
+  # Check MPI
   - which mpiexec
   # Check location
   - pwd
@@ -34,6 +32,7 @@ pip-installation:
     - *virtualenv
     - pip install -e .
     # Check result of installation
+    - python -c "import loadskernel"
     - which loads-kernel
     - which model-viewer
     - which loads-compare
@@ -86,10 +85,12 @@ Pytest:
     - *virtualenv
     # Install with -e (in editable mode) to allow the tracking of the test coverage
     - pip install -e .
+    - pip list
     # Get the examples repository
     - git clone https://gitlab-ci-token:${CI_JOB_TOKEN}@gitlab.dlr.de/loads-kernel/loads-kernel-examples.git
     # Run the actual testing of the code with pytest
-    - pytest -v --basetemp=./test_tmp --cov=loadskernel  --cov=modelviewer --cov=loadscompare --junitxml=testresult.xml
+    # Using python -m pytest is necessary because pytest has the habit of not looking in the site-packages of the venv
+    - python -m pytest -v --basetemp=./tmp --cov=loadskernel  --cov=modelviewer --cov=loadscompare --junitxml=testresult.xml
     # Create some reports
     - coverage report
     - coverage xml -o coverage.xml

.pytest.ini

@@ -0,0 +1,10 @@
+[pytest]
+filterwarnings =
+	# Filter to avoid annoying deprecation warnings raised by third party code
+	# Step 1: ignore all deprecation warnings
+    ignore::DeprecationWarning::
+    # Step 2: re-activate deprecation warnings for own modules
+    default::DeprecationWarning:loadskernel.*
+    default::DeprecationWarning:loadsviewer.*
+    default::DeprecationWarning:modelviewer.*
+    default::DeprecationWarning:tests.*

loadskernel/io_functions/read_bdf.py

@@ -190,7 +190,7 @@ def aggregate_cards(self, card_names):
         for card_name in card_names:
             old_size = self.cards[card_name].shape[0]
             sort_by_field = self.card_interpreters[card_name].field_names[0]
-            self.cards[card_name] = self.cards[card_name].groupby(by=sort_by_field, as_index=False).agg(sum)
+            self.cards[card_name] = self.cards[card_name].groupby(by=sort_by_field, as_index=False).agg("sum")
             new_size = self.cards[card_name].shape[0]
             if old_size != new_size:
                 logging.info('Aggregating {} {}s'.format(old_size - new_size, card_name))

loadskernel/io_functions/read_mona.py

@@ -416,7 +416,7 @@ def add_SET1(pandas_sets):
     set_values = []
     for _, row in pandas_sets[['values']].iterrows():
         # create a copy of the current row to work with
-        my_row = copy.deepcopy(row[0])
+        my_row = copy.deepcopy(row.iloc[0])
         # remove all None values
         my_row = [item for item in my_row if item is not None]
         values = []
@@ -448,8 +448,8 @@ def add_AECOMP(pandas_aecomps):
     # Loop over the rows to check for NaNs and None, which occur in case an empty field was in the list.
     # Then, select only the valid list items.
     for _, row in pandas_aecomps[['LISTID']].iterrows():
-        is_id = [pd.notna(x) for x in row[0]]
-        list_id.append(list(compress(row[0], is_id)))
+        is_id = [pd.notna(x) for x in row.iloc[0]]
+        list_id.append(list(compress(row.iloc[0], is_id)))
 
     aecomp = {}
     aecomp['name'] = pandas_aecomps['NAME'].to_list()

loadskernel/plotting_standard.py

@@ -166,7 +166,7 @@ def plot_forces_deformation_interactive(self):
 
     def plot_monstations(self, filename_pdf):
         # launch plotting
-        self.pp = PdfPages(filename_pdf)
+        self.pp = PdfPages(filename_pdf, keep_empty=False)
         self.potato_plots()
         if self.cuttingforces_wing:
             self.cuttingforces_along_wing_plots()
@@ -570,7 +570,7 @@ def plot_eigenvalues(self):
                 ax[1].minorticks_on()
                 ax[1].axis([-1.0, 1.0, imin, imax])
                 # connect with y-axis from left hand plot
-                ax[0].get_shared_y_axes().join(ax[0], ax[1])
+                ax[0].sharey(ax[1])
                 ax[1].yaxis.set_tick_params(which='both', labelleft=False, labelright=False)
                 ax[1].yaxis.offsetText.set_visible(False)
                 # add legend

loadskernel/spline_rules.py

@@ -32,7 +32,8 @@ def rules_point(grid_i, grid_d):
     # All dependent grids are mapped to one grid point, which might be CG or MAC
     # Assumption: the relevant point is expected to be the only/first point in the independet grid
     splinerules = {}
-    splinerules[int(grid_i['ID'])] = list(grid_d['ID'])
+    assert len(grid_i['ID']) == 1, "The independent grid 'grid_i' may have only one grid point for this kind of spline rules."
+    splinerules[grid_i['ID'][0]] = list(grid_d['ID'])
     return splinerules
 
 

tests/requirements.txt

@@ -0,0 +1,20 @@
+# These are the requirements to be used during regression testing
+mayavi
+traits
+traitsui
+pyface
+pyfmi
+h5py
+mpi4py
+pytest-cov
+psutil
+pytables
+pyyaml
+matplotlib
+mamba
+pandas
+openpyxl
+jupyter
+jupyter-book
+flake8
+pylint

tests/test_dummy.py

@@ -0,0 +1,7 @@
+import platform
+
+
+def test_dummy():
+    print('The dummy test is executed.')
+    print('Running on python version {}'.format(platform.python_version()))
+    pass

tests/test_gui.py

@@ -1,7 +1,10 @@
 import logging
 
-from loadscompare import compare
-from modelviewer import view
+try:
+    from loadscompare import compare
+    from modelviewer import view
+except ImportError:
+    pass
 
 
 class TestLoadsCompare():