From 4828b84d79c375639f4db462ef6da85c83adb364 Mon Sep 17 00:00:00 2001
From: foefl <f.foerster@d-opt.com>
Date: Wed, 26 Mar 2025 16:12:14 +0100
Subject: [PATCH] prepare tables

---
 src/delta_barth/constants.py |  2 ++
 src/delta_barth/databases.py | 61 ++++++++++++++++++++++++++++++++++++
 2 files changed, 63 insertions(+)
 create mode 100644 src/delta_barth/databases.py

diff --git a/src/delta_barth/constants.py b/src/delta_barth/constants.py
index 4a85a6f..b2a49cd 100644
--- a/src/delta_barth/constants.py
+++ b/src/delta_barth/constants.py
@@ -20,6 +20,8 @@ LOGGING_TO_FILE: Final[bool] = True
 LOGGING_TO_STDERR: Final[bool] = True
 LOG_FILENAME: Final[str] = "dopt-delbar.log"
 
+# ** databases
+DB_ECHO: Final[bool] = True
 
 # ** error handling
 DEFAULT_INTERNAL_ERR_CODE: Final[int] = 100
diff --git a/src/delta_barth/databases.py b/src/delta_barth/databases.py
new file mode 100644
index 0000000..56b5999
--- /dev/null
+++ b/src/delta_barth/databases.py
@@ -0,0 +1,61 @@
+from pathlib import Path
+
+import sqlalchemy as sql
+
+from delta_barth.constants import DB_ECHO
+
+# ** meta
+metadata = sql.MetaData()
+
+
+def get_engine(
+    db_path: Path,
+) -> sql.Engine:
+    path = db_path.resolve()
+    connection_str: str = f"sqlite:///{str(path)}"
+    engine = sql.create_engine(connection_str, echo=DB_ECHO)
+    return engine
+
+
+# ** table declarations
+# ** ---- common
+perf_meas = sql.Table(
+    "performance_measurement",
+    metadata,
+    sql.Column("id", sql.Integer, primary_key=True),
+    sql.Column("execution_duration", sql.Float),
+    sql.Column("pipeline_name", sql.String(length=30)),
+)
+# ** ---- forecasts
+sf_stats = sql.Table(
+    "sales_forecast_statistics",
+    metadata,
+    sql.Column("id", sql.Integer, primary_key=True),
+    sql.Column("error_code", sql.Integer),
+    sql.Column("error_msg", sql.String(length=200)),
+    sql.Column("length_dataset", sql.Integer),
+    sql.Column("score_mae", sql.Float, nullable=True),
+    sql.Column("score_r2", sql.Float, nullable=True),
+    sql.Column("best_start_year", sql.Integer, nullable=True),
+)
+sf_XGB = sql.Table(
+    "sales_forecast_XGB_parameters",
+    metadata,
+    sql.Column("id", sql.Integer, primary_key=True),
+    sql.Column(
+        "forecast_id",
+        sql.Integer,
+        sql.ForeignKey(
+            "sales_forecast_statistics.id", onupdate="CASCADE", ondelete="CASCADE"
+        ),
+        unique=True,
+    ),
+    sql.Column("n_estimators", sql.Integer),
+    sql.Column("learning_rate", sql.Float),
+    sql.Column("max_depth", sql.Integer),
+    sql.Column("min_child_weight", sql.Integer),
+    sql.Column("gamma", sql.Float),
+    sql.Column("subsample", sql.Float),
+    sql.Column("colsample_bytree", sql.Float),
+    sql.Column("early_stopping_rounds", sql.Integer),
+)