Modified .rst files and jupyter notebooks to improve readability

.github/workflows/pre-commit.yml

@@ -57,5 +57,13 @@ jobs:
       #     Rscript -e "install.packages('WGCNA', repos='https://cran.r-project.org')"
       #   shell: bash
 
+      - name: Cache pre-commit hooks
+        uses: actions/cache@v3
+        with:
+          path: ~/.cache/pre-commit
+          key: pre-commit-${{ runner.os }}-${{ hashFiles('.pre-commit-config.yaml') }}
+          restore-keys: pre-commit-${{ runner.os }}-
+
       - name: Run Pre-Commit Checks
+        if: matrix.os != 'ubuntu-latest'
         run: pre-commit run --all-files --show-diff-on-failure

.gitignore

@@ -25,6 +25,7 @@ doc_examples_outdated
 dpmon_output
 .enviroment
 TCGA-BRCA_Dataset_testing*.ipynb
+
 # Other example data and tests not needed in the repo.
 
 Output**

README.md

@@ -17,6 +17,10 @@ It provides tools for network construction, embedding generation, clustering, an
 
 ![BioNeuralNet Workflow](assets/BioNeuralNet.png)
 
+## Documentation
+
+**[BioNeuralNet Documentation & Examples](https://bioneuralnet.readthedocs.io/en/latest/)**
+
 ## Table of Contents
 
 - [1. Installation](#1-installation)

bioneuralnet/downstream_task/dpmon.py

@@ -246,7 +246,7 @@ def setup_device(gpu, cuda):
 def slice_omics_datasets(
     omics_dataset: pd.DataFrame, adjacency_matrix: pd.DataFrame
 ) -> List[pd.DataFrame]:
-    logger.info("Slicing omics dataset based on network nodes.")
+    logger.debug("Slicing omics dataset based on network nodes.")
     omics_network_nodes_names = adjacency_matrix.index.tolist()
 
     # Clean omics dataset columns
@@ -272,7 +272,7 @@ def build_omics_networks_tg(
     omics_datasets: List[pd.DataFrame],
     clinical_data: pd.DataFrame,
 ) -> List[Data]:
-    logger.info("Building PyTorch Geometric Data object from adjacency matrix.")
+    logger.debug("Building PyTorch Geometric Data object from adjacency matrix.")
     omics_network_nodes_names = adjacency_matrix.index.tolist()
 
     G = nx.from_pandas_adjacency(adjacency_matrix)
@@ -285,7 +285,7 @@ def build_omics_networks_tg(
 
     if clinical_data is not None and not clinical_data.empty:
         clinical_vars = clinical_data.columns.tolist()
-        logger.info(f"Using clinical vars for node features: {clinical_vars}")
+        logger.debug(f"Using clinical vars for node features: {clinical_vars}")
         omics_dataset = omics_datasets[0]
         missing_nodes = set(omics_network_nodes_names) - set(omics_dataset.columns)
         if missing_nodes:
@@ -385,12 +385,12 @@ def run_standard_training(dpmon_params, adjacency_matrix, combined_omics, clinic
         avg_accuracy = sum(accuracies) / len(accuracies)
         std_accuracy = statistics.stdev(accuracies) if len(accuracies) > 1 else 0.0
         logger.info(f"Best Accuracy: {best_accuracy:.4f}")
-        logger.info(f"Average Accuracy: {avg_accuracy:.4f}")
-        logger.info(f"Standard Deviation of Accuracy: {std_accuracy:.4f}")
+        logger.info(f"Average Accuracy across {len(accuracies)} models: {avg_accuracy:.4f}")
+        logger.info(f"Standard Deviation across all models: {std_accuracy:.4f}")
+        logger.info(f"Returning best model predictions and average accuracy (predictions, avg_accuracy).")
 
     return best_predictions_df, avg_accuracy
 
-
 def run_hyperparameter_tuning(
     dpmon_params, adjacency_matrix, combined_omics, clinical_data
 ):
@@ -408,7 +408,7 @@ def run_hyperparameter_tuning(
         "weight_decay": tune.loguniform(1e-4, 1e-1),
         "nn_hidden_dim1": tune.choice([4, 8, 16, 32, 64, 128]),
         "nn_hidden_dim2": tune.choice([4, 8, 16, 32, 64, 128]),
-        "num_epochs": tune.choice([2, 16, 64, 512, 1024, 4096, 8192]),
+        "num_epochs": tune.choice([16, 64, 256, 512, 1024,2048, 4096, 8192]),
     }
 
     reporter = CLIReporter(metric_columns=["loss", "accuracy", "training_iteration"])
@@ -501,7 +501,7 @@ def short_dirname_creator(trial):
 
         result = tune.run(
             tune_train_n,
-            resources_per_trial={"cpu": 2, "gpu": gpu_resources},
+            resources_per_trial={"cpu": 1, "gpu": gpu_resources},
             config=pipeline_configs,
             num_samples=10,
             verbose=0,
@@ -515,9 +515,7 @@ def short_dirname_creator(trial):
         best_trial = result.get_best_trial("loss", "min", "last")
         logger.info("Best trial config: {}".format(best_trial.config))
         logger.info("Best trial final loss: {}".format(best_trial.last_result["loss"]))
-        logger.info(
-            "Best trial final accuracy: {}".format(best_trial.last_result["accuracy"])
-        )
+        logger.info("Best trial final accuracy: {}".format(best_trial.last_result["accuracy"]))
         best_configs.append(best_trial.config)
 
     best_configs_df = pd.DataFrame(best_configs)
@@ -534,15 +532,13 @@ def train_model(model, criterion, optimizer, train_data, train_labels, epoch_num
         optimizer.step()
 
         if (epoch + 1) % 10 == 0 or epoch == 0:
-            logger.info(f"Epoch [{epoch+1}/{epoch_num}], Loss: {loss.item():.4f}")
+            logger.debug(f"Epoch [{epoch+1}/{epoch_num}], Loss: {loss.item():.4f}")
 
     model.eval()
     with torch.no_grad():
         predictions, _ = model(train_data, train_labels["omics_network"])
         _, predicted = torch.max(predictions, 1)
-        accuracy = (predicted == train_labels["labels"]).sum().item() / len(
-            train_labels["labels"]
-        )
+        accuracy = (predicted == train_labels["labels"]).sum().item() / len(train_labels["labels"])
         logger.info(f"Training Accuracy: {accuracy:.4f}")
 
     return accuracy

bioneuralnet/utils/preprocess.py

@@ -6,7 +6,6 @@
 from sklearn.ensemble import RandomForestClassifier, RandomForestRegressor
 from sklearn.feature_selection import f_classif, f_regression
 from statsmodels.stats.multitest import multipletests
-from typing import Callable, TypeAlias, overload
 
 from .logger import get_logger
 logger = get_logger(__name__)
@@ -61,7 +60,7 @@ def preprocess_clinical(X: pd.DataFrame, y: pd.Series, top_k: int = 10, scale: b
     else:
         df_cat_encoded = pd.DataFrame(index=df_numeric_scaled.index)
 
-    df_combined = pd.concat([df_numeric_scaled, df_cat_encoded, df_ignore],axis=1,join="inner")
+    df_combined = pd.concat([df_numeric_scaled, df_cat_encoded],axis=1,join="inner")
     df_features = df_combined.loc[:, df_combined.std(axis=0) > 0]
 
     if y_series.nunique() <= 10:
@@ -94,6 +93,12 @@ def preprocess_clinical(X: pd.DataFrame, y: pd.Series, top_k: int = 10, scale: b
     for idx in selected_idx:
         selected_columns.append(feature_names[idx])
 
+    df_selected = df_features[selected_columns].copy()
+
+    for col in df_ignore.columns:
+        if col in selected_columns:
+            df_selected[col] = df_ignore[col]
+
     return df_features[selected_columns]
 
 def clean_inf_nan(df: pd.DataFrame) -> pd.DataFrame: