refactor: modernize project and improve error handling and documentation

Dockerfile

@@ -0,0 +1,17 @@
+FROM python:3.11-slim-bookworm
+COPY --from=ghcr.io/astral-sh/uv:latest /uv /uvbin/uv
+
+WORKDIR /app
+
+COPY pyproject.toml uv.lock ./
+RUN /uvbin/uv pip install --system --requirement pyproject.toml
+
+COPY . .
+
+RUN mkdir -p output config .cache && chmod -R 777 output config .cache
+
+ENV PYTHONPATH=.
+ENV INTERVAL=3600
+ENV UV_CACHE_DIR=/app/.cache
+
+CMD ["/uvbin/uv", "run", "python", "main.py", "run"]

README.md

@@ -0,0 +1,64 @@
+# Strompreis & Netz Pipeline
+
+Automatisierte End-to-End Daten-Pipeline zur Erfassung, Analyse und Visualisierung von Strommarkt- und Wetterdaten.
+
+## 🏗 Architektur (Medallion Prinzip)
+
+Das Projekt folgt einer sauberen Data-Engineering-Struktur unter Verwendung von DuckDB und Polars:
+
+1.  **Bronze Layer (Raw):** Unveränderte API-Antworten von SMARD und Bright Sky werden als historische Fakten gespeichert.
+2.  **Gold Layer (Combined):** Transformierte, bereinigte und über Zeitstempel zusammengeführte Daten (Inner Join), die direkt für Analysen und das Dashboard bereitstehen.
+
+## 🚀 Features
+
+- **Datensammlung:** Automatisiertes Abrufen von:
+  - Strompreisen, Netzlast, Gesamterzeugung, Wind- & Solar-Erzeugung (SMARD API).
+  - Umfassenden Wetterdaten (Temperatur, Wind, Solarstrahlung, etc. via Bright Sky API).
+- **Dashboard:** Interaktive Visualisierung mit Streamlit (Preise, Energiemix, Wetter-Details).
+- **REST-API:** FastAPI-Schnittstelle für den programmatischen Zugriff auf aktuelle Daten.
+- **Orchestrierung:** Vollständige Docker-Integration inklusive eines automatisierten Schedulers.
+
+## 🛠 Installation & Betrieb
+
+### Mit Docker (Empfohlen)
+
+Das gesamte System (Pipeline, API, Dashboard) wird mit einem Befehl gestartet:
+
+```bash
+docker-compose up --build
+```
+- **Dashboard:** `http://localhost:8501`
+- **API:** `http://localhost:8000/docs`
+
+### Lokal (Entwicklung)
+
+1. **Abhängigkeiten installieren:**
+   ```bash
+   uv sync
+   ```
+
+2. **Pipeline manuell ausführen:**
+   ```bash
+   uv run python main.py run
+   ```
+
+3. **Dashboard starten:**
+   ```bash
+   uv run streamlit run dashboard/app.py
+   ```
+
+## ⚙️ Konfiguration
+
+Die Konfiguration erfolgt über `config/config.yaml` oder Umgebungsvariablen mit dem Präfix `STROM_`:
+- `STROM_SMARD__REGION`: Region-Code (Standard: DE-LU)
+- `STROM_DATABASE__PATH`: Pfad zur DuckDB Datei
+- `INTERVAL`: Ausführungsintervall des Schedulers im Docker-Container (in Sekunden).
+
+## 🧪 Tests
+
+```bash
+uv run pytest
+```
+
+## Lizenz
+MIT

api/main.py

@@ -0,0 +1,51 @@
+"""
+REST API for accessing processed electricity price and network data.
+"""
+
+import sys
+from pathlib import Path
+from fastapi import FastAPI, HTTPException
+
+# Add project root to sys.path
+project_root = str(Path(__file__).parent.parent)
+if project_root not in sys.path:
+    sys.path.append(project_root)
+
+from utils import database as db
+from utils.config_loader import settings
+
+app = FastAPI(
+    title="Strompreis API",
+    description="Provides access to cleaned and joined electricity price, network, and weather data."
+)
+
+@app.get("/current-price")
+async def get_current_price():
+    """
+    Returns the most recent price record from the Gold (combined) layer.
+    """
+    try:
+        with db.get_connection() as con:
+            res = con.execute(
+                "SELECT timestamp, price FROM combined ORDER BY timestamp DESC LIMIT 1"
+            ).fetchone()
+            
+            if not res:
+                raise HTTPException(status_code=404, detail="No data available")
+            
+            return {
+                "timestamp": res[0],
+                "price_eur_mwh": res[1],
+                "price_ct_kwh": round(res[1] / 10, 2)
+            }
+    except Exception as e:
+        raise HTTPException(status_code=500, detail=f"Database error: {str(e)}")
+
+@app.get("/status")
+async def get_status():
+    """Returns basic system configuration and status."""
+    return {
+        "region": settings.smard.region,
+        "database": settings.database.path,
+        "status": "operational"
+    }

collectors/pricing.py

@@ -1,29 +0,0 @@
-from datetime import datetime
-import time
-import polars as pl
-from utils import request_utils
-
-
-def fetch_prices(region: str = "DE-LU", filter_id: int = 4169) -> pl.DataFrame:
-    current_time = 0
-    closest = 0
-    now_ms = int(time.time()) * 1000
-    index_url = (
-        f"https://www.smard.de/app/chart_data/{filter_id}/{region}/index_hour.json"
-    )
-    timestamps = request_utils.make_requests(index_url, timeout=20)
-    closest = min(timestamps.get("timestamps", []), key=lambda x: abs(x - now_ms))
-
-    current_time = closest
-    url = f"https://www.smard.de/app/chart_data/{filter_id}/{region}/{filter_id}_{region}_hour_{current_time}.json"
-    data = request_utils.make_requests(url, timeout=20)
-    series = data.get("series", [])
-    clean_series = [row for row in series if row[1] is not None]
-
-    df = pl.DataFrame(
-        {
-            "timestamp": [row[0] for row in clean_series],
-            "price": [row[1] for row in clean_series],
-        }
-    )
-    return df

collectors/smard.py

@@ -0,0 +1,58 @@
+"""
+Collector for SMARD (Electricity Market Data) API.
+"""
+
+import time
+import logging
+import polars as pl
+from utils import request_utils
+from utils.config_loader import settings
+
+logger = logging.getLogger(__name__)
+
+def fetch_smard_data(filter_id: int, region: str = None) -> pl.DataFrame:
+    """
+    Fetches time-series data from the SMARD API for a specific filter.
+
+    Args:
+        filter_id: The SMARD data identifier (e.g., 4169 for price).
+        region: The region code (defaults to config value).
+
+    Returns:
+        A Polars DataFrame with 'timestamp' (ms) and 'value'.
+    """
+    region = region or settings.smard.region
+    base_url = settings.smard.base_url
+
+    now_ms = int(time.time()) * 1000
+    index_url = f"{base_url}/{filter_id}/{region}/index_hour.json"
+    
+    try:
+        index_data = request_utils.make_requests(index_url)
+        if not index_data or "timestamps" not in index_data:
+            logger.warning(f"No index data for filter {filter_id}")
+            return pl.DataFrame()
+            
+        closest_ts = min(index_data["timestamps"], key=lambda x: abs(x - now_ms))
+        
+        data_url = (
+            f"{base_url}/{filter_id}/{region}/"
+            f"{filter_id}_{region}_hour_{closest_ts}.json"
+        )
+        payload = request_utils.make_requests(data_url)
+        series = payload.get("series", [])
+        
+        clean_series = [row for row in series if row[1] is not None]
+        if not clean_series:
+            return pl.DataFrame()
+
+        return pl.DataFrame(
+            {
+                "timestamp": [row[0] for row in clean_series],
+                "value": [row[1] for row in clean_series],
+            },
+            schema={"timestamp": pl.Int64, "value": pl.Float64}
+        )
+    except Exception as e:
+        logger.error(f"Failed to fetch SMARD filter {filter_id}: {e}")
+        return pl.DataFrame()

collectors/weather.py

@@ -1,11 +1,30 @@
+"""
+Collector for Bright Sky (Weather) API.
+"""
+
+import logging
 from datetime import datetime, timedelta, timezone
-from utils import request_utils
+
 import polars as pl
 
+from utils import request_utils
+from utils.config_loader import settings
+
+logger = logging.getLogger(__name__)
+
+
+def fetch_weather(lat: float = None, lon: float = None) -> pl.DataFrame:
+    """
+    Fetches historical weather data from Bright Sky API.
+    Filters for the specific columns defined in the Bronze schema.
+
+    Returns:
+        A Polars DataFrame containing filtered raw weather parameters.
+    """
+    lat = lat or settings.brightsky.lat
+    lon = lon or settings.brightsky.lon
+    url = settings.brightsky.base_url
 
-def fetch_weather(lat: float = 52.52, lon: float = 13.41) -> pl.DataFrame:
-    url = "https://api.brightsky.dev/weather"
-    headers = {"Accept": "application/json"}
     current_utc = datetime.now(timezone.utc)
     payload = {
         "date": (current_utc - timedelta(hours=72)).isoformat(),
@@ -15,8 +34,9 @@ def fetch_weather(lat: float = 52.52, lon: float = 13.41) -> pl.DataFrame:
         "units": "dwd",
         "tz": "Etc/UTC",
     }
-    data = request_utils.make_requests(url, headers, payload, 20)
-    relevant_fields = [
+
+    schema_cols = [
+        "timestamp",
         "temperature",
         "wind_speed",
         "solar",
@@ -24,13 +44,20 @@ def fetch_weather(lat: float = 52.52, lon: float = 13.41) -> pl.DataFrame:
         "cloud_cover",
         "precipitation",
     ]
-    critical_fields = ["temperature", "wind_speed"]
 
-    weather_data = [
-        {k: v for k, v in entry.items() if k in relevant_fields or k in ["timestamp"]}
-        for entry in data["weather"]
-        if all(entry.get(field) is not None for field in critical_fields)
-    ]
+    try:
+        data = request_utils.make_requests(url, params=payload)
+        weather_list = data.get("weather", [])
 
-    df_weather = pl.DataFrame(weather_data)
-    return df_weather
+        if not weather_list:
+            logger.warning("No weather data returned from API.")
+            return pl.DataFrame()
+
+        df = pl.DataFrame(weather_list)
+
+        available_cols = [c for c in schema_cols if c in df.columns]
+        return df.select(available_cols)
+
+    except Exception as e:
+        logger.error(f"Failed to fetch weather data: {e}")
+        return pl.DataFrame()

config/config.yaml

@@ -0,0 +1,12 @@
+smard:
+  region: "DE-LU"
+  filter_id: 4169
+  base_url: "https://www.smard.de/app/chart_data"
+
+brightsky:
+  lat: 52.52
+  lon: 13.41
+  base_url: "https://api.brightsky.dev/weather"
+
+database:
+  path: "output/pipeline.duckdb"

dashboard/app.py

@@ -0,0 +1,73 @@
+import sys
+from pathlib import Path
+
+import duckdb
+import polars as pl
+import streamlit as st
+
+project_root = str(Path(__file__).parent.parent)
+if project_root not in sys.path:
+    sys.path.append(project_root)
+
+from utils.config_loader import settings
+
+st.set_page_config(page_title="Strompreis & Netz Dashboard", layout="wide")
+
+st.title("⚡ Strompreis, Netz & Wetter Dashboard")
+
+
+@st.cache_data(ttl=300)
+def load_data():
+    try:
+        con = duckdb.connect(settings.database.path)
+        df = con.execute("SELECT * FROM combined ORDER BY timestamp DESC").pl()
+        con.close()
+        return df.fill_null(0)
+    except Exception as e:
+        st.error(f"Fehler beim Laden der Daten: {e}")
+        return pl.DataFrame()
+
+
+df = load_data()
+
+if df.is_empty():
+    st.warning("Keine Daten gefunden. Bitte Pipeline starten.")
+else:
+    latest = df.head(1).to_dicts()[0]
+    m1, m2, m3, m4 = st.columns(4)
+    m1.metric("Preis", f"{latest['price']:.2f} €/MWh")
+    m2.metric("Netzlast", f"{latest['load_forecast'] / 1000:.1f} GW")
+    m3.metric("Erneuerbare", f"{(latest['wind_total'] + latest['pv']) / 1000:.1f} GW")
+    m4.metric("Außentemp.", f"{latest['temperature']:.1f} °C")
+
+    tab1, tab2, tab3 = st.tabs(["Markt & Netz", "Wetter-Details", "Rohdaten"])
+
+    with tab1:
+        st.subheader("Strompreis vs. Netzlast")
+        pdf = df.to_pandas()
+        st.line_chart(pdf.set_index("timestamp")[["price", "load_forecast"]])
+
+        st.subheader("Erzeugung Mix (Wind & Solar)")
+        st.area_chart(pdf.set_index("timestamp")[["wind_total", "pv"]])
+
+    with tab2:
+        st.subheader("Detaillierte Wetterdaten")
+        w_col1, w_col2 = st.columns(2)
+        with w_col1:
+            st.info("Temperaturverlauf")
+            st.line_chart(pdf.set_index("timestamp")["temperature"])
+        with w_col2:
+            st.info("Windgeschwindigkeit")
+            st.line_chart(pdf.set_index("timestamp")["wind_speed"])
+
+        st.info("Weitere Wetterfaktoren")
+        available_weather = [
+            c
+            for c in ["solar", "sunshine", "cloud_cover", "precipitation"]
+            if c in df.columns
+        ]
+        if available_weather:
+            st.line_chart(pdf.set_index("timestamp")[available_weather])
+
+    with tab3:
+        st.dataframe(df)

docker-compose.yaml

@@ -0,0 +1,36 @@
+services:
+  pipeline:
+    build: .
+    user: "1000:1000"
+    volumes:
+      - ./output:/app/output:z
+      - ./config:/app/config:z
+    command: ["/bin/bash", "scripts/scheduler.sh"]
+    environment:
+      - INTERVAL=3600
+    restart: unless-stopped
+
+  api:
+    build: .
+    user: "1000:1000"
+    ports:
+      - "8000:8000"
+    volumes:
+      - ./output:/app/output:z
+      - ./config:/app/config:z
+    command: ["python", "-m", "uvicorn", "api.main:app", "--host", "0.0.0.0", "--port", "8000"]
+    restart: unless-stopped
+
+  dashboard:
+    build: .
+    user: "1000:1000"
+    ports:
+      - "8501:8501"
+    volumes:
+      - ./output:/app/output:z
+      - ./config:/app/config:z
+    environment:
+      - STREAMLIT_BROWSER_GATHER_USAGE_STATS=false
+      - STREAMLIT_USAGE_STATS_ENABLED=false
+    command: ["streamlit", "run", "dashboard/app.py", "--server.port", "8501", "--server.address", "0.0.0.0"]
+    restart: unless-stopped

main.py

@@ -1,10 +1,18 @@
+"""
+Main entry point for the Strompreis Pipeline CLI.
+Coordinates data collection, transformation, and storage.
+"""
+
 import sys
-from collectors import pricing, weather
+import logging
+import click
+import polars as pl
+from collectors import smard, weather
 from transformators import transformator
 from utils import database as db
-import logging
-
+from utils.config_loader import settings
 
+# Structured logging configuration
 logging.basicConfig(
     level=logging.INFO,
     format="%(asctime)s - %(name)s - %(levelname)s - %(message)s",
@@ -12,39 +20,100 @@ logging.basicConfig(
 logger = logging.getLogger(__name__)
 
 
-def main():
-    logger.info("Starte Strompreis-Pipeline")
+@click.group()
+def cli():
+    """Strompreis Pipeline CLI Tools."""
+    pass
 
-    logger.info("Hole Preisdaten...")
-    df_prices_raw = pricing.fetch_prices()
-    logger.info(f"Preise: {len(df_prices_raw)} Zeilen")
 
-    logger.info("Hole Wetterdaten...")
-    df_weather_raw = weather.fetch_weather()
-    logger.info(f"Wetter: {len(df_weather_raw)} Zeilen")
+@cli.command()
+@click.option("--region", help="SMARD region code (e.g., DE-LU).")
+@click.option("--lat", type=float, help="Latitude for weather data.")
+@click.option("--lon", type=float, help="Longitude for weather data.")
+def run(region, lat, lon):
+    """
+    Executes the full ETL pipeline:
+    1. Bronze Layer: Fetch and store raw data.
+    2. Gold Layer: Transform and join data into a clean business view.
+    """
+    logger.info("Initializing Pipeline...")
 
-    logger.info("Transformiere Daten...")
-    df_prices_clean = transformator.transform_prices(df_prices_raw)
-    df_weather_clean = transformator.transform_weather(df_weather_raw)
-    df_joined = transformator.join_dataframes(df_prices_clean, df_weather_clean)
-    logger.info(f"Joined: {len(df_joined)} Zeilen")
+    try:
+        logger.info("Step 1/4: Fetching raw data from APIs...")
+        smard_queries = {
+            "prices_raw": settings.smard.price_filter,
+            "load_forecast_raw": settings.smard.load_forecast_filter,
+            "gen_total_raw": settings.smard.generation_total_filter,
+            "wind_onshore_raw": settings.smard.wind_onshore_filter,
+            "wind_offshore_raw": settings.smard.wind_offshore_filter,
+            "pv_raw": settings.smard.pv_filter,
+        }
 
-    logger.info("Speichere in DuckDB...")
+        smard_dfs = {
+            tbl: smard.fetch_smard_data(fid, region)
+            for tbl, fid in smard_queries.items()
+        }
+        df_weather_raw = weather.fetch_weather(lat, lon)
+
+        logger.info("Step 2/4: Persisting Bronze Layer...")
         with db.get_connection() as con:
             db.init_tables(con)
-        db.upsert_prices(con, df_prices_raw)
-        db.upsert_prices_clean(con, df_prices_clean)
-        db.upsert_weather(con, df_weather_raw)
-        db.upsert_combined(con, df_joined)
-        logger.info("Erfolgreich zur Datenbank hinzugefügt.")
+            for table, df in smard_dfs.items():
+                db.upsert_raw(con, table, df)
+            db.upsert_raw(con, "weather_raw", df_weather_raw)
+
+        logger.info("Step 3/4: Transforming and joining data (Gold Layer)...")
+
+        def to_gold_smard(df, name):
+            """Internal helper to convert raw SMARD df to gold format."""
+            if df.is_empty():
+                return pl.DataFrame(
+                    schema={"timestamp": pl.Datetime("ms", "UTC"), name: pl.Float64}
+                )
+            return df.with_columns(
+                pl.col("timestamp").cast(pl.Datetime("ms")).dt.replace_time_zone("UTC")
+            ).rename({"value": name})
+
+        df_p = to_gold_smard(smard_dfs["prices_raw"], "price")
+        df_l = to_gold_smard(smard_dfs["load_forecast_raw"], "load_forecast")
+        df_g = to_gold_smard(smard_dfs["gen_total_raw"], "generation_total")
+        df_w_on = to_gold_smard(smard_dfs["wind_onshore_raw"], "wind_on")
+        df_w_off = to_gold_smard(smard_dfs["wind_offshore_raw"], "wind_off")
+        df_pv = to_gold_smard(smard_dfs["pv_raw"], "pv")
+
+        df_wind = df_w_on.join(
+            df_w_off, on="timestamp", how="full", coalesce=True
+        ).fill_null(0)
+        df_wind = df_wind.with_columns(
+            (pl.col("wind_on") + pl.col("wind_off")).alias("wind_total")
+        )
+
+        df_weather_gold = transformator.transform_weather(df_weather_raw)
+
+        df_smard_all = (
+            df_p.join(df_l, on="timestamp", how="full", coalesce=True)
+            .join(df_g, on="timestamp", how="full", coalesce=True)
+            .join(
+                df_wind[["timestamp", "wind_total"]],
+                on="timestamp",
+                how="full",
+                coalesce=True,
+            )
+            .join(df_pv, on="timestamp", how="full", coalesce=True)
+        )
+
+        df_gold = df_smard_all.join(df_weather_gold, on="timestamp", how="inner")
+
+        logger.info("Step 4/4: Persisting Gold Layer...")
+        with db.get_connection() as con:
+            db.upsert_combined(con, df_gold)
+
+        logger.info("Pipeline completed successfully.")
+
+    except Exception as e:
+        logger.error(f"Pipeline execution failed: {e}", exc_info=True)
+        sys.exit(1)
 
 
 if __name__ == "__main__":
-    try:
-        main()
-    except KeyboardInterrupt:
-        logger.info("Pipeline durch Benutzer gestoppt")
-        sys.exit(0)
-    except Exception as e:
-        logger.error(f"Pipeline fehlgeschlagen: {e}", exc_info=True)
-        sys.exit(1)
+    cli()

pyproject.toml

@@ -5,8 +5,36 @@ description = "Add your description here"
 readme = "README.md"
 requires-python = ">=3.11"
 dependencies = [
+    "click>=8.3.1",
     "duckdb>=1.4.4",
+    "fastapi>=0.128.7",
     "polars>=1.38.1",
     "pyarrow>=23.0.0",
+    "pydantic>=2.12.5",
+    "pydantic-settings>=2.12.0",
+    "pyyaml>=6.0.3",
     "requests>=2.32.5",
+    "streamlit>=1.54.0",
+    "tenacity>=9.1.4",
+    "uvicorn>=0.40.0",
+]
+
+[tool.ruff]
+line-length = 88
+target-version = "py311"
+
+[tool.ruff.lint]
+select = ["E", "F", "I", "W", "C90", "B"]
+ignore = []
+
+[tool.pytest.ini_options]
+pythonpath = ["."]
+testpaths = ["tests"]
+
+[dependency-groups]
+dev = [
+    "pytest>=9.0.2",
+    "pytest-mock>=3.15.1",
+    "requests-mock>=1.12.1",
+    "ruff>=0.15.0",
 ]

scripts/scheduler.sh

@@ -0,0 +1,8 @@
+#!/bin/bash
+echo "Starte Pipeline Scheduler (Intervall: $INTERVAL Sekunden)"
+while true; do
+  echo "Führe Pipeline aus: $(date)"
+  python main.py run
+  echo "Pipeline beendet. Warte $INTERVAL Sekunden..."
+  sleep ${INTERVAL:-3600}
+done

transformators/transformator.py

@@ -1,76 +1,40 @@
+"""
+Data transformation logic for price and weather data.
+"""
+
 from datetime import timedelta
 import polars as pl
 
+def transform_weather(df: pl.DataFrame) -> pl.DataFrame:
+    """
+    Cleans and standardizes raw weather data.
+    """
+    if df.is_empty():
+        return pl.DataFrame()
+
+    relevant_cols = [
+        "timestamp", "temperature", "wind_speed", "solar", 
+        "sunshine", "cloud_cover", "precipitation"
+    ]
+    
+    # Filter for existing relevant columns
+    cols_to_keep = [c for c in relevant_cols if c in df.columns]
+    
+    return df.select(cols_to_keep).with_columns(
+        pl.col("timestamp")
+        .str.to_datetime(format="%Y-%m-%dT%H:%M:%S%z", time_zone="UTC")
+        .cast(pl.Datetime("ms", time_zone="UTC"))
+    ).sort("timestamp")
 
 def transform_prices(df: pl.DataFrame) -> pl.DataFrame:
+    """
+    Transforms raw SMARD timestamps to UTC Datetime.
+    """
+    if df.is_empty():
+        return pl.DataFrame()
+        
     return df.with_columns(
-        [
         pl.col("timestamp")
         .cast(pl.Datetime("ms"))
         .dt.replace_time_zone("UTC")
-            .alias("timestamp")
-        ]
-    )
-
-
-def transform_weather(df: pl.DataFrame) -> pl.DataFrame:
-    return df.with_columns(
-        [
-            pl.col("timestamp")
-            .str.to_datetime(
-                format="%Y-%m-%dT%H:%M:%S%z",
-                time_zone="UTC",
-            )
-            .cast(pl.Datetime("ms", time_zone="UTC"))
-            .alias("timestamp")
-        ]
-    )
-
-
-def join_dataframes(df_prices: pl.DataFrame, df_weather: pl.DataFrame) -> pl.DataFrame:
-    return df_prices.join(df_weather, on="timestamp", how="inner").select(
-        [
-            pl.col("timestamp"),
-            pl.col("price"),
-            pl.col(
-                [
-                    "temperature",
-                    "wind_speed",
-                    "solar",
-                    "sunshine",
-                    "cloud_cover",
-                    "precipitation",
-                ]
-            ),
-        ]
-    )
-
-
-#
-#
-# def join_dataframes(df_prices: pl.DataFrame, df_weather: pl.DataFrame) -> pl.DataFrame:
-#     return (
-#         df_prices.sort("timestamp")  # ← Pflicht!
-#         .join_asof(
-#             df_weather.sort("timestamp"),  # ← Pflicht!
-#             on="timestamp",
-#             strategy="nearest",  # ← nearest statt backward!
-#             tolerance=timedelta(hours=1),
-#         )
-#         .select(
-#             [
-#                 pl.col("timestamp"),
-#                 pl.col("price"),
-#                 pl.col(
-#                     [
-#                         "temperature",
-#                         "wind_speed",
-#                         "solar",
-#                         "sunshine",
-#                         "cloud_cover",
-#                         "precipitation",
-#                     ]
-#                 ),
-#             ]
-#         )
-#     )
+    ).sort("timestamp")

utils/config_loader.py

@@ -0,0 +1,54 @@
+from pathlib import Path
+
+import yaml
+from pydantic import BaseModel
+from pydantic_settings import BaseSettings, SettingsConfigDict
+
+
+class SmardConfig(BaseModel):
+    region: str = "DE-LU"
+    price_filter: int = 4169
+    load_forecast_filter: int = 4382
+    generation_total_filter: int = 122
+    wind_onshore_filter: int = 4069
+    wind_offshore_filter: int = 4068
+    pv_filter: int = 4070
+    base_url: str = "https://www.smard.de/app/chart_data"
+
+
+class BrightSkyConfig(BaseModel):
+    lat: float = 52.52
+    lon: float = 13.41
+    base_url: str = "https://api.brightsky.dev/weather"
+
+
+class DatabaseConfig(BaseModel):
+    path: str = "output/pipeline.duckdb"
+
+
+class Settings(BaseSettings):
+    model_config = SettingsConfigDict(
+        env_file=".env",
+        env_nested_delimiter="__",
+        env_prefix="STROM_",
+        extra="ignore"
+    )
+
+    smard: SmardConfig = SmardConfig()
+    brightsky: BrightSkyConfig = BrightSkyConfig()
+    database: DatabaseConfig = DatabaseConfig()
+
+
+def load_config(config_path: str = "config/config.yaml") -> Settings:
+    path = Path(config_path)
+    if not path.exists():
+        return Settings()
+
+    with open(path, "r") as f:
+        config_data = yaml.safe_load(f) or {}
+
+    return Settings(**config_data)
+
+
+# Global settings instance
+settings = load_config()

utils/database.py

@@ -1,10 +1,17 @@
+"""
+DuckDB database interface for Bronze (Raw) and Gold (Combined) layers.
+"""
+
 import duckdb
 import polars as pl
 from contextlib import contextmanager
+from utils.config_loader import settings
 
 
 @contextmanager
-def get_connection(db_path: str = "output/pipeline.duckdb"):
+def get_connection(db_path: str = None):
+    """Context manager for DuckDB connections."""
+    db_path = db_path or settings.database.path
     con = duckdb.connect(db_path)
     try:
         yield con
@@ -13,23 +20,26 @@ def get_connection(db_path: str = "output/pipeline.duckdb"):
 
 
 def init_tables(con: duckdb.DuckDBPyConnection):
-    con.execute("""
-    CREATE TABLE IF NOT EXISTS prices_raw (
-        timestamp BIGINT NOT NULL UNIQUE,
-        price DOUBLE NOT NULL
+    """
+    Initializes the database schema following the Medallion architecture.
+    """
+    # BRONZE LAYER (Raw API responses)
+    smard_tables = [
+        "prices_raw",
+        "load_forecast_raw",
+        "gen_total_raw",
+        "wind_onshore_raw",
+        "wind_offshore_raw",
+        "pv_raw",
+    ]
+    for table in smard_tables:
+        con.execute(
+            f"CREATE TABLE IF NOT EXISTS {table} (timestamp BIGINT PRIMARY KEY, value DOUBLE)"
         )
-    """)
-
-    con.execute("""
-    CREATE TABLE IF NOT EXISTS prices_clean (
-        timestamp TIMESTAMP NOT NULL UNIQUE,
-        price DOUBLE NOT NULL
-    )
-    """)
 
     con.execute("""
     CREATE TABLE IF NOT EXISTS weather_raw (
-        timestamp TIMESTAMP NOT NULL UNIQUE,
+        timestamp VARCHAR PRIMARY KEY,
         temperature DOUBLE,
         wind_speed DOUBLE,
         solar DOUBLE,
@@ -39,10 +49,15 @@ def init_tables(con: duckdb.DuckDBPyConnection):
     )
     """)
 
+    # GOLD LAYER (Transformed and Joined Business Data)
     con.execute("""
     CREATE TABLE IF NOT EXISTS combined (
-        timestamp TIMESTAMP NOT NULL UNIQUE,
-        price DOUBLE NOT NULL,
+        timestamp TIMESTAMP PRIMARY KEY,
+        price DOUBLE,
+        load_forecast DOUBLE,
+        generation_total DOUBLE,
+        wind_total DOUBLE,
+        pv DOUBLE,
         temperature DOUBLE,
         wind_speed DOUBLE,
         solar DOUBLE,
@@ -53,25 +68,23 @@ def init_tables(con: duckdb.DuckDBPyConnection):
     """)
 
 
-def upsert_prices(con: duckdb.DuckDBPyConnection, df: pl.DataFrame):
+def upsert_raw(con: duckdb.DuckDBPyConnection, table_name: str, df: pl.DataFrame):
+    """Inserts raw data using explicit columns to match the target table schema."""
+    if df.is_empty():
+        return
+
+    cols = con.execute(f"DESCRIBE {table_name}").pl()["column_name"].to_list()
+    df_to_insert = df.select([c for c in cols if c in df.columns])
+
     con.execute(
-        """INSERT INTO prices_raw SELECT * FROM df ON CONFLICT (timestamp) DO NOTHING;"""
-    )
-
-
-def upsert_prices_clean(con: duckdb.DuckDBPyConnection, df: pl.DataFrame):
-    con.execute(
-        """INSERT INTO prices_clean SELECT * FROM df ON CONFLICT (timestamp) DO NOTHING;"""
-    )
-
-
-def upsert_weather(con: duckdb.DuckDBPyConnection, df: pl.DataFrame):
-    con.execute(
-        """INSERT INTO weather_raw SELECT * FROM df ON CONFLICT (timestamp) DO NOTHING;"""
+        f"INSERT INTO {table_name} SELECT * FROM df_to_insert ON CONFLICT (timestamp) DO NOTHING;"
     )
 
 
 def upsert_combined(con: duckdb.DuckDBPyConnection, df: pl.DataFrame):
+    """Inserts business-ready data into the Gold layer."""
+    if df.is_empty():
+        return
     con.execute(
-        """INSERT INTO combined SELECT * FROM df ON CONFLICT (timestamp) DO NOTHING;"""
+        "INSERT INTO combined SELECT * FROM df ON CONFLICT (timestamp) DO NOTHING;"
     )

utils/request_utils.py

@@ -1,23 +1,41 @@
+import logging
+
 import requests
+from tenacity import (
+    retry,
+    retry_if_exception_type,
+    stop_after_attempt,
+    wait_exponential,
+)
+
+logger = logging.getLogger(__name__)
 
 
+@retry(
+    stop=stop_after_attempt(3),
+    wait=wait_exponential(multiplier=1, min=2, max=10),
+    retry=retry_if_exception_type((requests.exceptions.RequestException,)),
+    reraise=True,
+)
 def make_requests(
-    url: str, headers: dict = {}, payload: dict = {}, timeout: int = 20
+    url: str, headers: dict = None, params: dict = None, timeout: int = 20
 ) -> dict:
+    headers = headers or {}
+    params = params or {}
     try:
-        res = requests.get(url, headers=headers, params=payload, timeout=timeout)
+        logger.debug(f"Requesting URL: {url} with params: {params}")
+        res = requests.get(url, headers=headers, params=params, timeout=timeout)
         res.raise_for_status()
         return res.json()
     except requests.exceptions.HTTPError as errh:
-        print("HTTP Error")
-        print(errh.args[0])
-        raise errh
+        logger.error(f"HTTP Error: {errh}")
+        raise
     except requests.ConnectionError as errc:
-        print("http connection Error")
-        raise errc
+        logger.error(f"Connection Error: {errc}")
+        raise
     except requests.exceptions.Timeout as errt:
-        print("http connection timeout")
-        raise errt
+        logger.error(f"Timeout Error: {errt}")
+        raise
     except Exception as e:
-        print("unknown exception")
-        raise e
+        logger.error(f"Unknown Exception: {e}")
+        raise