Update GitHub Pages dependencies: specify exact versions for improved…

.github/workflows/gh-pages.yml

@@ -27,6 +27,8 @@ jobs:
     steps:
       - name: Checkout
         uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
       - name: Setup Python
         uses: actions/setup-python@v5
         with:
@@ -38,8 +40,12 @@ jobs:
       - name: Install dependencies
         run: |
           python -m pip install --upgrade pip
-          pip install mkdocs-material mkdocstrings[python] mkdocs-git-revision-date-localized-plugin mkdocs-plugin-tags mike
+          pip install "mkdocs>=1.5" "mkdocs-material>=9.5"
+          pip install "mkdocstrings>=0.25" "mkdocstrings-python>=1.9" "mike>=2.0" "mkdocs-autorefs>=1.0" "mkdocs-git-revision-date-localized-plugin>=1.2"
+      #          pip install Markdown==3.10 Jinja2==3.1.6 pymdown-extensions==10.18 mkdocs==1.6.1 mkdocstrings==1.0.0 mkdocstrings-python==2.0.1 mike==2.1.3 mkdocs-autorefs==1.4.3 mkdocs-material==8.2.7 mkdocs-git-revision-date-localized-plugin==1.5.0 mkdocs-plugin-tags==1.0.2
       - name: Build site
+        env:
+          PYTHONPATH: ${{ github.workspace }}
         run: mkdocs build --strict
       - name: Upload artifact
         uses: actions/upload-pages-artifact@v3

docs/api/uqdd.md

@@ -7,4 +7,9 @@ tags:
 
 # uqdd package
 
+The top-level package overview is provided here. For detailed modules, see:
+- [Core Utils](../api/uqdd-utils.md)
+- [Data modules](../api/uqdd-data.md)
+- [Model modules](../api/uqdd-models.md)
+
 ::: uqdd

docs/changelog.md

@@ -17,3 +17,5 @@ tags:
 - Configured GitHub Pages deployment workflow (`.github/workflows/gh-pages.yml`).
 - Enabled versioning support via `mike` plugin (optional).
 - Improved local strict builds by enabling `git-revision-date-localized` only on CI.
+
+See also the [tags](tags.md) page for topic-based navigation.

docs/index.md

@@ -13,9 +13,12 @@ Welcome to the UQDD documentation. This site covers installation, data pipeline,
 [Combining Bayesian and Evidential Uncertainty Quantification for Improved Bioactivity Modeling](https://pubs.acs.org/doi/10.1021/acs.jcim.5c01597)
 
 ## Highlights
+
 - Hybrid UQ models: EOE (Ensemble of Evidential) and EMC (Evidential MC Dropout)
 - Baselines: Deep ensembles, MC-Dropout, and probabilistic neural networks
 - Papyrus++ datasets for xC50 and Kx bioactivity endpoints
 - Comprehensive evaluation: performance, calibration, probabilistic scoring, decision utility
 
+See the [tags](tags.md) page for topics across the documentation.
+
 Use the top navigation to get started.

docs/reference/mind-map.md

@@ -9,6 +9,6 @@ tags:
 
 The following mind map provides a high-level overview of the paper and the UQDD project:
 
-![Mind Map](assets/MindMap.png)
+![Mind Map](../assets/MindMap.png)
 
 For details, refer to the specific sections throughout the documentation.

mkdocs.yml

@@ -66,7 +66,7 @@ plugins:
   - mkdocstrings:
       handlers:
         python:
-          paths: ["uqdd"]
+          paths: ['.']
           options:
             show_source: false
             docstring_style: numpy
@@ -79,10 +79,9 @@ plugins:
             show_root_heading: true
             show_signature: true
             heading_level: 2
-            show_submodules: true
+            show_submodules: false
   - mike
   - git-revision-date-localized:
-      enabled_if_env: CI
       fallback_to_build_date: true
       enable_creation_date: true
       enable_git_follow: false

uqdd/models/pnn.py

@@ -259,9 +259,6 @@ def run_pnn(config: Optional[dict] = None) -> nn.Module:  # uq: bool = False
     config : Optional[dict], optional
         Configuration dictionary for model training, by default None.
 
-    uq : bool
-        Whether to use aleatoric uncertainty estimation, by default
-
     Returns
     -------
     nn.Module
@@ -274,33 +271,6 @@ def run_pnn(config: Optional[dict] = None) -> nn.Module:  # uq: bool = False
         logger=LOGGER,
     )
 
-    # if uq:
-    #     dataloaders = get_dataloader(config, device=DEVICE, logger=LOGGER)
-    #     preds, labels, alea_vars = predict(
-    #         best_model, dataloaders["test"], device=DEVICE
-    #     )
-    #     # Then comes the predict metrics part
-    #     metrics, plots, uct_logger = evaluate_predictions(
-    #         config, preds, labels, alea_vars, "pnn", LOGGER
-    #     )
-    #     # RECALIBRATION # Get Calibration / Validation Set
-    #     preds_val, labels_val, alea_vars_val = predict(
-    #         best_model, dataloaders["val"], device=DEVICE
-    #     )
-    #     iso_recal_model = recalibrate_model(
-    #         preds_val,
-    #         labels_val,
-    #         alea_vars_val,
-    #         preds,
-    #         labels,
-    #         alea_vars,
-    #         config=config,
-    #         uct_logger=uct_logger,
-    #     )
-    #     uct_logger.wandb_log()
-    #     wandb.finish()
-    #     return best_model, iso_recal_model, metrics, plots
-
     return best_model