Upload src/preprocess.py with huggingface_hub

src/preprocess.py

@@ -0,0 +1,52 @@
+import pandas as pd
+
+
+REQUIRED_COLUMNS = [
+    "engine_rpm",
+    "coolant_temperature",
+    "oil_pressure",
+    "fuel_pressure",
+    "intake_temp",
+    "battery_voltage",
+]
+
+
+def validate_input(df: pd.DataFrame) -> None:
+    """Check if all required columns are present"""
+    missing_cols = [col for col in REQUIRED_COLUMNS if col not in df.columns]
+    if missing_cols:
+        raise ValueError(f"Missing required columns: {missing_cols}")
+
+
+def handle_missing_values(df: pd.DataFrame) -> pd.DataFrame:
+    """Simple missing value handling"""
+    return df.fillna(df.median(numeric_only=True))
+
+
+def feature_engineering(df: pd.DataFrame) -> pd.DataFrame:
+    """Create derived features (if used in training)"""
+
+    
+    df["temp_pressure_ratio"] = df["coolant_temperature"] / (df["oil_pressure"] + 1)
+    df["rpm_pressure_interaction"] = df["engine_rpm"] * df["oil_pressure"]
+
+    return df
+
+
+def ensure_column_order(df: pd.DataFrame) -> pd.DataFrame:
+    """Ensure correct feature order for model"""
+    return df[sorted(df.columns)]
+
+
+def preprocess_input(df: pd.DataFrame) -> pd.DataFrame:
+    """Full preprocessing pipeline"""
+
+    validate_input(df)
+
+    df = handle_missing_values(df)
+
+    df = feature_engineering(df)
+
+    df = ensure_column_order(df)
+
+    return df