Remove ReadTheDocs configuration and update documentation for clarity and consistency

Author: karl-koschutnig

.readthedocs.yml

@@ -44,7 +44,7 @@ What the installer does (high-level):
 
 ```bash
 source braingraph_pipeline/bin/activate
-python scripts/validate_setup.py --config configs/braingraph_default_config.json
+python scripts/validate_setup.py --config configs/          braingraph_default_config.json
 ```
 
 ---

docs/cli_steps.md

@@ -24,7 +24,7 @@ This method serves as a rigorous baseline and validation tool.
 
 ### Design
 -   **Split-Half Validation**: The cohort is split into two "waves" (Wave 1 and Wave 2).
--   **Exhaustive Search**: A grid of parameters is defined (e.g., FA $\in \{0.1, 0.2\}$, Angle $\in \{30, 60\}$).
+-   **Exhaustive Search**: A grid of parameters is defined (e.g., FA $\in \\{0.1, 0.2\\}$, Angle $\in \\{30, 60\\}$).
 -   **Evaluation**: Every combination is run on both waves.
 -   **Selection**: We select parameters that:
     1.  Score highly in Wave 1.
@@ -37,7 +37,12 @@ This method is computationally expensive ($O(N^k)$ where $k$ is the number of pa
 
 The objective function maximizes a composite score derived from graph-theoretic metrics:
 
-$$ Score = w_1 \cdot Density_{score} + w_2 \cdot Efficiency_{score} + w_3 \cdot SmallWorld_{score} ... $$
+$$
+\mathrm{Score} =
+w_1 \cdot \mathrm{Density}_{\mathrm{score}} +
+w_2 \cdot \mathrm{Efficiency}_{\mathrm{score}} +
+w_3 \cdot \mathrm{SmallWorld}_{\mathrm{score}} + \cdots
+$$
 
 -   **Density**: Penalizes unconnected or fully connected graphs.
 -   **Global Efficiency**: Measures integration.

docs/methods.md

@@ -0,0 +1,20 @@
+/*
+Disable colored syntax highlighting in code blocks.
+Works with/without Pygments or other highlighters.
+*/
+
+pre code,
+.codehilite,
+.codehilite pre,
+.highlight,
+.highlight pre {
+  color: inherit;
+}
+
+.codehilite span,
+.highlight span {
+  color: inherit !important;
+  background: none !important;
+  font-weight: inherit !important;
+  font-style: inherit !important;
+}

docs/rt_d_integration.md

@@ -4,47 +4,55 @@ All OptiConn workflows follow the same core pattern: **parameter space explorati
 
 ## Bayesian optimization → Apply (recommended)
 1. Find optimal parameters on a pilot subset:
-   ```bash
-   python opticonn.py tune-bayes \
-     -i /path/to/pilot_data \
-     -o studies/bayes_opt \
-     --config configs/braingraph_default_config.json \
-     --modalities qa \
-     --n-iterations 30 \
-     --sample-subjects
-   ```
+
+```console
+python opticonn.py tune-bayes \
+    -i /path/to/pilot_data \
+    -o studies/bayes_opt \
+    --config configsbraingraph_default_config.json \
+    --modalities qa \
+    --n-iterations 30 \
+    --sample-subjects
+```
 2. Inspect selection:
-   ```bash
-   python opticonn.py select -i studies/bayes_opt --modality qa
-   ```
+
+```console
+python opticonn.py select -i studies/bayes_opt --modality qa   
+```
+
 3. Apply to the full dataset:
-   ```bash
-   python opticonn.py apply \
-     -i /path/to/full_dataset \
-     --optimal-config studies/bayes_opt/qa/bayesian_optimization_results.json \
-     -o studies/final_analysis
-   ```
+
+```console
+python opticonn.py apply \
+    -i /path/to/full_dataset \
+    --optimal-config studies/bayes_oqa/bayesian_optimization_results.json \
+    -o studies/final_analysis
+```
 
 ## Cross-validation bootstrap (with Bayes seeding)
 1. Run Bayesian optimization (small pilot) as above.
 2. Seed cross-validation with the Bayes result:
-   ```bash
-   python scripts/cross_validation_bootstrap_optimizer.py \
-     -i /path/to/pilot_data \
-     -o studies/cv \
-     --extraction-config configs/demo_config.json \
-     --from-bayes studies/bayes_opt/qa/bayesian_optimization_results.json \
-     --subjects 3 \
-     --max-parallel 1 \
-     --verbose
-   ```
-   - Uses two waves by default; metrics/atlases stay fixed from the base config.
-   - Seeded parameters come from `best_parameters` in the Bayes results.
+
+```sh
+python scriptcross_validation_bootstrap_optimizepy \
+    -i /path/to/pilot_data \
+    -o studies/cv \
+    --extraction-config configdemo_config.json \
+    --from-bayes studies/bayes_optqbayesian_optimization_results.json \
+    --subjects 3 \
+    --max-parallel 1 \
+    --verbose
+```
+
+- Uses two waves by default; metrics/atlases stay fixed from the base config.
+- Seeded parameters come from `best_parameters` in the Bayes results.
 
 ## Apply-only with known optimal config
-```bash
+
+```sh
 python opticonn.py apply \
-  -i /path/to/full_dataset \
-  --optimal-config studies/bayes_opt/qa/bayesian_optimization_results.json \
-  -o studies/final_analysis
+    -i /path/to/full_dataset \
+    --optimal-config studies/bayes_opt/qa/ \ bayesian_optimization_results.json \
+    -o studies/final_analysis
 ```
+

docs/stylesheets/no_code_colors.css

@@ -11,7 +11,6 @@ nav:
   - Workflows: workflows.md
   - CLI & Steps: cli_steps.md
   - Demos: demos.md
-  - RTD Integration: rt_d_integration.md
   - Configuration: configuration.md
   - Validation Notes: validation.md
   - User Guide: user_guide.md
@@ -21,7 +20,15 @@ theme:
   name: readthedocs
 markdown_extensions:
   - admonition
-  - codehilite
+  - fenced_code
+  - pymdownx.arithmatex:
+      generic: true
   - toc:
       permalink: true
   - tables
+
+extra_css:
+  - stylesheets/no_code_colors.css
+
+extra_javascript:
+  - https://cdn.jsdelivr.net/npm/mathjax@3/es5/tex-mml-chtml.js

docs/workflows.md

@@ -0,0 +1 @@
+# Test package for OptiConn scripts