Merge pull request #226 from JDACS4C-IMPROVE/rajeeja/parsl_csa_githubactions

o GraphDRP parsl csa workflow

Author: rajeeja

.github/workflows/README.md

@@ -0,0 +1,23 @@
+# IMPROVE UnitTest GitHub Actions (unittest-actions.yml)
+
+This repository contains a GitHub Actions workflow to run unit tests for the IMPROVE library. The workflow is triggered on every push to the repository.
+
+The workflow is defined in the `.github/workflows/unittest-actions.yml` file. Here are the steps it performs:
+
+1. **Checkout Code**: The repository code is checked out using the `actions/checkout` action.
+2. **Install and Check**: The workflow installs the IMPROVE library in editable mode and verifies the installation by printing the version. It then sets the `PYTHONPATH` environment variable and creates necessary directories for data. Finally, it runs the `test_drp_params.py` test script.
+
+# Docker Parsl GraphDRP CSA Workflow (docker_gdrp_parsl.yml)
+
+This GitHub Actions workflow is designed to automate the testing and execution of the GraphDRP CSA (Cross-Species Analysis) workflow using Docker. The workflow performs the following steps:
+
+1. **Checkout the Repository**: The workflow checks out the IMPROVE repository code.
+2. **Install Missing Libraries**: Installs necessary system libraries such as `libxrender1`, `libxext6`, and `tree`.
+3. **Install and Check**: Verifies the Python and Conda environments, activates the `GraphDRP` Conda environment (already has csa-data downloaded), and installs the required Python packages.
+4. **Run Preprocessing and CSA Workflows**: Executes the preprocessing and Parsl CSA workflows using the specified configuration files.
+## Configuration Files
+
+The workflow uses the following configuration files:
+
+- **`parsl_csa_githubactions.ini`**: This configuration file is used to set up the parameters for the Parsl CSA workflow. It includes settings for data paths, model parameters, and other relevant configurations.
+

.github/workflows/docker_gdrp_parsl.yml

@@ -0,0 +1,51 @@
+name: Docker Parsl GraphDRP CSA workflow
+on: [push]
+jobs:
+  test:
+    runs-on: ubuntu-latest
+    container:
+      image: rajeeja/graphdrp-image:latest
+    steps:
+      - uses: actions/checkout@v4
+      - name: Install missing libraries
+        run: |
+          apt-get update && apt-get install -y libxrender1 libxext6 tree #added tree here
+      - name: Install and check
+        run: |
+          which python
+          python --version
+          which conda
+          conda env list
+          eval "$(conda shell.bash hook)"
+          conda activate graphdrp
+          pip install -e .
+          export PYTHONPATH=$PYTHONPATH:.
+          python -c "from improvelib.applications.drug_response_prediction.config import DRPPreprocessConfig; print('Import successful')"
+          BASE_DIR=$(pwd)
+          ls -ltrh
+          cd /app/GraphDRP
+          ls -ltrh
+          conda list -n graphdrp
+          python -c "import sys; print(sys.path)"
+          cd $BASE_DIR/workflows/csa/parsl/
+          pip install cryptography==37.0.1
+          pip install pyopenssl==22.0.0
+          pip install cffi==1.14.5
+          pip install parsl numpy pandas scikit-learn pyyaml
+          mv /app/GraphDRP/csa_data .
+          echo "Preprocess started"
+          ls -ltrh
+          python workflow_preprocess.py --config_file ./parsl_csa_githubactions.ini
+          echo "Preprocess done"
+          python workflow_csa.py --config_file ./parsl_csa_githubactions.ini
+          echo "Parsl CSA done"
+          echo "Tree of ./parsl_csa_exp3:"
+          tree ./parsl_csa_exp3
+          echo "Content of ./parsl_csa_exp3/infer/CCLE-CCLE/split_0/test_scores.json:"
+          cat ./parsl_csa_exp3/infer/CCLE-CCLE/split_0/test_scores.json
+          echo "Content of ./parsl_csa_exp3/infer/CCLE-CCLE/split_1/test_scores.json:"
+          cat ./parsl_csa_exp3/infer/CCLE-CCLE/split_1/test_scores.json
+          echo "Content of ./parsl_csa_exp3/infer/CCLE-gCSI/split_0/test_scores.json:"
+          cat ./parsl_csa_exp3/infer/CCLE-gCSI/split_0/test_scores.json
+          echo "Content of ./parsl_csa_exp3/infer/gCSI-CCLE/split_0/test_scores.json:"
+          cat ./parsl_csa_exp3/infer/gCSI-CCLE/split_0/test_scores.json

workflows/csa/parsl/parsl_csa_githubactions.ini

@@ -0,0 +1,24 @@
+[DEFAULT]
+input_dir = ./csa_data/raw_data
+y_col_name = auc
+use_singularity = False
+singularity_image = path_to_your_singularity_container.sif
+hyperparameters_file = ./hyperparameters_default.json
+model_name = graphdrp
+model_scripts_dir = /app/GraphDRP
+model_environment = graphdrp
+only_cross_study = False
+epochs = 2
+input_supp_data_dir = ''
+
+; # Exp 3
+output_dir = ./parsl_csa_exp3
+source_datasets = ["CCLE","gCSI"]
+target_datasets = ["CCLE","gCSI"]
+split = ["0","1"]
+
+[Preprocess]
+
+[Train]
+
+[Infer]