pre-commit

Author: SimonBlanke

src/surfaces/_surrogates/__init__.py

@@ -17,6 +17,13 @@
 - list_ml_surrogates: List registered functions and status
 """
 
+from ._ml_surrogate_trainer import (
+    MLSurrogateTrainer,
+    list_ml_surrogates,
+    train_all_ml_surrogates,
+    train_missing_ml_surrogates,
+    train_ml_surrogate,
+)
 from ._surrogate_loader import (
     SurrogateLoader,
     get_surrogate_path,
@@ -29,13 +36,6 @@
 from ._surrogate_validator import (
     SurrogateValidator,
 )
-from ._ml_surrogate_trainer import (
-    MLSurrogateTrainer,
-    train_ml_surrogate,
-    train_all_ml_surrogates,
-    train_missing_ml_surrogates,
-    list_ml_surrogates,
-)
 
 __all__ = [
     # Loader

src/surfaces/_surrogates/_dashboard/__init__.py

@@ -44,7 +44,9 @@ def run_dashboard():
         print("Install it with: pip install surfaces[dashboard]")
         sys.exit(1)
 
-    subprocess.run([sys.executable, "-m", "streamlit", "run", str(app_path), "--server.headless", "true"])
+    subprocess.run(
+        [sys.executable, "-m", "streamlit", "run", str(app_path), "--server.headless", "true"]
+    )
 
 
 __all__ = ["run_dashboard"]

src/surfaces/_surrogates/_dashboard/_pages/details.py

@@ -6,7 +6,6 @@
 Details page - Deep dive into a single function.
 """
 
-
 import pandas as pd
 import streamlit as st
 
@@ -57,12 +56,14 @@ def render():
         st.warning(f"No surrogate model for {selected}")
 
     # Tabs for different info
-    tab1, tab2, tab3, tab4 = st.tabs([
-        "Metadata",
-        "Parameters",
-        "Training History",
-        "Validation History",
-    ])
+    tab1, tab2, tab3, tab4 = st.tabs(
+        [
+            "Metadata",
+            "Parameters",
+            "Training History",
+            "Validation History",
+        ]
+    )
 
     with tab1:
         render_metadata(surrogate)
@@ -95,9 +96,21 @@ def render_metadata(surrogate: dict):
         if surrogate["has_surrogate"]:
             st.write(f"- Training Samples: {surrogate['n_samples'] or 'N/A'}")
             st.write(f"- Invalid Samples: {surrogate['n_invalid_samples'] or 0}")
-            st.write(f"- Training R2: {surrogate['training_r2']:.4f}" if surrogate['training_r2'] else "- Training R2: N/A")
-            st.write(f"- Training MSE: {surrogate['training_mse']:.6f}" if surrogate['training_mse'] else "- Training MSE: N/A")
-            st.write(f"- Training Time: {surrogate['training_time_sec']:.1f}s" if surrogate['training_time_sec'] else "- Training Time: N/A")
+            st.write(
+                f"- Training R2: {surrogate['training_r2']:.4f}"
+                if surrogate["training_r2"]
+                else "- Training R2: N/A"
+            )
+            st.write(
+                f"- Training MSE: {surrogate['training_mse']:.6f}"
+                if surrogate["training_mse"]
+                else "- Training MSE: N/A"
+            )
+            st.write(
+                f"- Training Time: {surrogate['training_time_sec']:.1f}s"
+                if surrogate["training_time_sec"]
+                else "- Training Time: N/A"
+            )
         else:
             st.write("No training data available.")
 
@@ -115,8 +128,12 @@ def render_metadata(surrogate: dict):
 
     with col2:
         st.write("**Tracking Info**")
-        st.write(f"- Last Synced: {surrogate['last_synced_at'][:19] if surrogate['last_synced_at'] else 'Never'}")
-        st.write(f"- Created: {surrogate['created_at'][:19] if surrogate['created_at'] else 'Unknown'}")
+        st.write(
+            f"- Last Synced: {surrogate['last_synced_at'][:19] if surrogate['last_synced_at'] else 'Never'}"
+        )
+        st.write(
+            f"- Created: {surrogate['created_at'][:19] if surrogate['created_at'] else 'Unknown'}"
+        )
         if surrogate["onnx_file_hash"]:
             st.write(f"- File Hash: `{surrogate['onnx_file_hash'][:16]}...`")
 
@@ -148,11 +165,13 @@ def render_parameters(surrogate: dict):
             param_type = "numeric"
             values_str = "-"
 
-        param_data.append({
-            "Parameter": name,
-            "Type": param_type,
-            "Values": values_str,
-        })
+        param_data.append(
+            {
+                "Parameter": name,
+                "Type": param_type,
+                "Values": values_str,
+            }
+        )
 
     df = pd.DataFrame(param_data)
     st.dataframe(df, use_container_width=True, hide_index=True)
@@ -180,20 +199,25 @@ def render_training_history(function_name: str):
             # Simple duration calculation
             try:
                 from datetime import datetime
+
                 start = datetime.fromisoformat(job["started_at"])
                 end = datetime.fromisoformat(job["completed_at"])
                 dur_sec = (end - start).total_seconds()
                 duration = f"{dur_sec:.1f}s"
             except Exception:
                 pass
 
-        df_data.append({
-            "Started": job["started_at"][:19] if job["started_at"] else "-",
-            "Duration": duration,
-            "Status": job["status"],
-            "Triggered By": job["triggered_by"],
-            "Error": job["error_message"][:30] + "..." if job["error_message"] and len(job["error_message"]) > 30 else (job["error_message"] or "-"),
-        })
+        df_data.append(
+            {
+                "Started": job["started_at"][:19] if job["started_at"] else "-",
+                "Duration": duration,
+                "Status": job["status"],
+                "Triggered By": job["triggered_by"],
+                "Error": job["error_message"][:30] + "..."
+                if job["error_message"] and len(job["error_message"]) > 30
+                else (job["error_message"] or "-"),
+            }
+        )
 
     df = pd.DataFrame(df_data)
 
@@ -228,10 +252,12 @@ def render_validation_history(function_name: str):
         chart_data = []
         for run in reversed(runs):  # Oldest first for chart
             if run["r2_score"] is not None:
-                chart_data.append({
-                    "Date": run["validated_at"][:10] if run["validated_at"] else "",
-                    "R2": run["r2_score"],
-                })
+                chart_data.append(
+                    {
+                        "Date": run["validated_at"][:10] if run["validated_at"] else "",
+                        "R2": run["r2_score"],
+                    }
+                )
 
         if chart_data:
             chart_df = pd.DataFrame(chart_data)
@@ -242,16 +268,18 @@ def render_validation_history(function_name: str):
 
     df_data = []
     for run in runs:
-        df_data.append({
-            "Date": run["validated_at"][:19] if run["validated_at"] else "-",
-            "Type": run["validation_type"],
-            "Samples": str(run["n_samples"]) if run["n_samples"] else "-",
-            "R2": f"{run['r2_score']:.4f}" if run["r2_score"] else "-",
-            "MAE": f"{run['mae']:.4f}" if run["mae"] else "-",
-            "RMSE": f"{run['rmse']:.4f}" if run["rmse"] else "-",
-            "Max Error": f"{run['max_error']:.4f}" if run["max_error"] else "-",
-            "Speedup": f"{run['speedup_factor']:.0f}x" if run["speedup_factor"] else "-",
-        })
+        df_data.append(
+            {
+                "Date": run["validated_at"][:19] if run["validated_at"] else "-",
+                "Type": run["validation_type"],
+                "Samples": str(run["n_samples"]) if run["n_samples"] else "-",
+                "R2": f"{run['r2_score']:.4f}" if run["r2_score"] else "-",
+                "MAE": f"{run['mae']:.4f}" if run["mae"] else "-",
+                "RMSE": f"{run['rmse']:.4f}" if run["rmse"] else "-",
+                "Max Error": f"{run['max_error']:.4f}" if run["max_error"] else "-",
+                "Speedup": f"{run['speedup_factor']:.0f}x" if run["speedup_factor"] else "-",
+            }
+        )
 
     df = pd.DataFrame(df_data)
 

src/surfaces/_surrogates/_dashboard/_pages/overview.py

@@ -91,9 +91,13 @@ def render():
         "Function Name": lambda x: x["function_name"],
         "R2 Score": lambda x: x["latest_r2"] or 0,
         "Samples": lambda x: x["n_samples"] or 0,
-        "Status": lambda x: ["Good", "Needs Attention", "Not Validated", "Missing"].index(x["status"]),
+        "Status": lambda x: ["Good", "Needs Attention", "Not Validated", "Missing"].index(
+            x["status"]
+        ),
     }
-    filtered = sorted(filtered, key=sort_key_map[sort_by], reverse=(sort_by == "R2 Score" or sort_by == "Samples"))
+    filtered = sorted(
+        filtered, key=sort_key_map[sort_by], reverse=(sort_by == "R2 Score" or sort_by == "Samples")
+    )
 
     # Summary metrics
     st.divider()
@@ -117,15 +121,17 @@ def render():
     # Build dataframe for display
     df_data = []
     for row in filtered:
-        df_data.append({
-            "Function": row["function_name"],
-            "Type": row["function_type"],
-            "Has Surrogate": "Yes" if row["has_surrogate"] else "No",
-            "Samples": str(row["n_samples"]) if row["n_samples"] else "-",
-            "Training R2": f"{row['training_r2']:.4f}" if row["training_r2"] else "-",
-            "Validation R2": f"{row['latest_r2']:.4f}" if row["latest_r2"] else "-",
-            "Status": row["status"],
-        })
+        df_data.append(
+            {
+                "Function": row["function_name"],
+                "Type": row["function_type"],
+                "Has Surrogate": "Yes" if row["has_surrogate"] else "No",
+                "Samples": str(row["n_samples"]) if row["n_samples"] else "-",
+                "Training R2": f"{row['training_r2']:.4f}" if row["training_r2"] else "-",
+                "Validation R2": f"{row['latest_r2']:.4f}" if row["latest_r2"] else "-",
+                "Status": row["status"],
+            }
+        )
 
     df = pd.DataFrame(df_data)
 
@@ -136,10 +142,7 @@ def highlight_status(val):
 
     # Display table
     if len(df) > 0:
-        styled_df = df.style.applymap(
-            highlight_status,
-            subset=["Status"]
-        )
+        styled_df = df.style.applymap(highlight_status, subset=["Status"])
         st.dataframe(
             styled_df,
             use_container_width=True,

src/surfaces/_surrogates/_dashboard/_pages/training.py

@@ -122,7 +122,9 @@ def progress_callback(msg: str):
         )
 
     with col2:
-        train_single = st.button("Train", use_container_width=True, type="primary", key="training_train_btn")
+        train_single = st.button(
+            "Train", use_container_width=True, type="primary", key="training_train_btn"
+        )
 
     if train_single and selected_function:
         st.divider()
@@ -162,14 +164,18 @@ def progress_callback(msg: str):
     if jobs:
         df_data = []
         for job in jobs:
-            df_data.append({
-                "Function": job["function_name"],
-                "Started": job["started_at"][:19] if job["started_at"] else "-",
-                "Completed": job["completed_at"][:19] if job["completed_at"] else "-",
-                "Status": job["status"],
-                "Triggered By": job["triggered_by"],
-                "Error": job["error_message"][:50] + "..." if job["error_message"] and len(job["error_message"]) > 50 else (job["error_message"] or "-"),
-            })
+            df_data.append(
+                {
+                    "Function": job["function_name"],
+                    "Started": job["started_at"][:19] if job["started_at"] else "-",
+                    "Completed": job["completed_at"][:19] if job["completed_at"] else "-",
+                    "Status": job["status"],
+                    "Triggered By": job["triggered_by"],
+                    "Error": job["error_message"][:50] + "..."
+                    if job["error_message"] and len(job["error_message"]) > 50
+                    else (job["error_message"] or "-"),
+                }
+            )
 
         df = pd.DataFrame(df_data)
 

src/surfaces/_surrogates/_dashboard/_pages/validation.py

@@ -87,9 +87,7 @@ def progress_callback(msg: str):
                 st.text("\n".join(logs[-20:]))
 
         with st.spinner("Validating all surrogates..."):
-            results = validate_all(
-                validation_type, n_samples, random_seed, progress_callback
-            )
+            results = validate_all(validation_type, n_samples, random_seed, progress_callback)
 
         success_count = sum(1 for r in results if r["success"])
         fail_count = len(results) - success_count
@@ -118,7 +116,9 @@ def progress_callback(msg: str):
         )
 
     with col2:
-        validate_single = st.button("Validate", use_container_width=True, type="primary", key="validation_validate_btn")
+        validate_single = st.button(
+            "Validate", use_container_width=True, type="primary", key="validation_validate_btn"
+        )
 
     if validate_single and selected_function:
         st.divider()
@@ -167,10 +167,12 @@ def progress_callback(msg: str):
                 y_surr = data["y_surrogate"]
 
                 # Create scatter plot data
-                scatter_df = pd.DataFrame({
-                    "Actual": y_real,
-                    "Predicted": y_surr,
-                })
+                scatter_df = pd.DataFrame(
+                    {
+                        "Actual": y_real,
+                        "Predicted": y_surr,
+                    }
+                )
 
                 # Plot
                 try:
@@ -188,8 +190,10 @@ def progress_callback(msg: str):
                     max_val = max(y_real.max(), y_surr.max())
                     fig.add_shape(
                         type="line",
-                        x0=min_val, y0=min_val,
-                        x1=max_val, y1=max_val,
+                        x0=min_val,
+                        y0=min_val,
+                        x1=max_val,
+                        y1=max_val,
                         line=dict(color="red", dash="dash"),
                     )
 
@@ -238,15 +242,17 @@ def progress_callback(msg: str):
     if runs:
         df_data = []
         for run in runs:
-            df_data.append({
-                "Function": run["function_name"],
-                "Type": run["validation_type"],
-                "Samples": str(run["n_samples"]) if run["n_samples"] else "-",
-                "R2": f"{run['r2_score']:.4f}" if run["r2_score"] else "-",
-                "MAE": f"{run['mae']:.4f}" if run["mae"] else "-",
-                "Speedup": f"{run['speedup_factor']:.0f}x" if run["speedup_factor"] else "-",
-                "Date": run["validated_at"][:19] if run["validated_at"] else "-",
-            })
+            df_data.append(
+                {
+                    "Function": run["function_name"],
+                    "Type": run["validation_type"],
+                    "Samples": str(run["n_samples"]) if run["n_samples"] else "-",
+                    "R2": f"{run['r2_score']:.4f}" if run["r2_score"] else "-",
+                    "MAE": f"{run['mae']:.4f}" if run["mae"] else "-",
+                    "Speedup": f"{run['speedup_factor']:.0f}x" if run["speedup_factor"] else "-",
+                    "Date": run["validated_at"][:19] if run["validated_at"] else "-",
+                }
+            )
 
         df = pd.DataFrame(df_data)