From 128133761f7259dc19feb29c2823ceb641023d17 Mon Sep 17 00:00:00 2001 From: Meryemel-gham Date: Tue, 15 Sep 2026 16:16:18 +0200 Subject: [PATCH] feat(apps): cree les six tables data et l'hypertable readings --- .../e6d2026091501_create_data_schema.py | 216 ++++++++++++++++++ apps/backend/app/models/__init__.py | 4 + apps/backend/app/models/energy.py | 209 +++++++++++++++++ 3 files changed, 429 insertions(+) create mode 100644 apps/backend/alembic/versions/e6d2026091501_create_data_schema.py create mode 100644 apps/backend/app/models/energy.py diff --git a/apps/backend/alembic/versions/e6d2026091501_create_data_schema.py b/apps/backend/alembic/versions/e6d2026091501_create_data_schema.py new file mode 100644 index 0000000..87146d6 --- /dev/null +++ b/apps/backend/alembic/versions/e6d2026091501_create_data_schema.py @@ -0,0 +1,216 @@ +"""Création des six tables Data et de l'hypertable readings. + +Revision ID: e6d2026091501 +Revises: 5353c0e4f094 +""" + +from alembic import op +import sqlalchemy as sa +from sqlalchemy.dialects import postgresql + +revision = "e6d2026091501" +down_revision = "5353c0e4f094" +branch_labels = None +depends_on = None + + +def upgrade() -> None: + # ### commands auto generated by Alembic - please adjust! ### + op.create_table( + "datasets", + sa.Column("dataset_id", sa.BigInteger(), autoincrement=True, nullable=False), + sa.Column("dataset_name", sa.Text(), nullable=False), + sa.Column("archive_sha256", sa.String(length=64), nullable=False), + sa.Column("storage_uri", sa.Text(), nullable=False), + sa.Column("source_timezone", sa.Text(), nullable=True), + sa.Column( + "metadata", postgresql.JSONB(none_as_null=True, astext_type=sa.Text()), nullable=False + ), + sa.CheckConstraint("dataset_id > 0", name="ck_datasets_positive_id"), + sa.PrimaryKeyConstraint("dataset_id"), + sa.UniqueConstraint("archive_sha256", name="uq_datasets_archive_sha256"), + ) + op.create_table( + "sites", + sa.Column("site_id", sa.Text(), nullable=False), + sa.Column("site_name", sa.Text(), nullable=False), + sa.Column("site_type", sa.Text(), nullable=False), + sa.Column("location", sa.Text(), nullable=True), + sa.Column("capacity_kw", sa.Double(), nullable=True), + sa.Column("status", sa.Text(), nullable=True), + sa.PrimaryKeyConstraint("site_id"), + ) + op.create_table( + "predictions", + sa.Column("prediction_id", sa.BigInteger(), autoincrement=True, nullable=False), + sa.Column("site_id", sa.Text(), nullable=False), + sa.Column( + "created_at", + sa.DateTime(timezone=True), + server_default=sa.text("now()"), + nullable=False, + ), + sa.Column("target_at", sa.DateTime(timezone=True), nullable=False), + sa.Column("target_metric", sa.Text(), nullable=False), + sa.Column("period_minutes", sa.Integer(), nullable=True), + sa.Column("predicted_value", sa.Double(), nullable=True), + sa.Column("model_reference", sa.Text(), nullable=False), + sa.Column("status", sa.Text(), nullable=False), + sa.Column("failure_reason", sa.Text(), nullable=True), + sa.CheckConstraint( + "(status = 'available' AND predicted_value IS NOT NULL AND failure_reason IS NULL) OR (status IN ('insufficient_data', 'error') AND predicted_value IS NULL AND failure_reason IS NOT NULL)", + name="ck_predictions_status", + ), + sa.CheckConstraint( + "target_metric <> 'consumption_kwh' OR period_minutes IS NOT NULL", + name="ck_predictions_energy_period", + ), + sa.CheckConstraint( + "target_metric IN ('consumption_kwh', 'consumption_kw')", name="ck_predictions_metric" + ), + sa.CheckConstraint( + "period_minutes IS NULL OR period_minutes > 0", name="ck_predictions_period" + ), + sa.ForeignKeyConstraint( + ["site_id"], ["sites.site_id"], name="fk_predictions_site", ondelete="RESTRICT" + ), + sa.PrimaryKeyConstraint("prediction_id"), + sa.UniqueConstraint("prediction_id", "site_id", name="uq_predictions_id_site"), + ) + op.create_index( + "ix_predictions_site_target", "predictions", ["site_id", "target_at"], unique=False + ) + op.create_table( + "readings", + sa.Column("reading_id", sa.BigInteger(), autoincrement=True, nullable=False), + sa.Column("site_id", sa.Text(), nullable=False), + sa.Column("timestamp", sa.DateTime(timezone=True), nullable=False), + sa.Column("source", sa.Text(), nullable=False), + sa.Column("dataset_id", sa.BigInteger(), nullable=True), + sa.Column("consumption_kw", sa.Double(), nullable=True), + sa.Column("consumption_kwh", sa.Double(), nullable=True), + sa.Column("consumption_euros", sa.Numeric(precision=14, scale=2), nullable=True), + sa.Column("voltage_v", sa.Double(), nullable=True), + sa.Column("current_a", sa.Double(), nullable=True), + sa.Column("power_factor", sa.Double(), nullable=True), + sa.Column("temperature_celsius", sa.Double(), nullable=True), + sa.Column("humidity_percent", sa.Double(), nullable=True), + sa.Column("solar_irradiance_wm2", sa.Double(), nullable=True), + sa.Column("is_working_hours", sa.Boolean(), nullable=True), + sa.Column("data_quality", sa.Text(), nullable=True), + sa.Column("null_reasons", postgresql.ARRAY(sa.Text()), nullable=True), + sa.Column( + "imputed_values", postgresql.JSONB(none_as_null=True, astext_type=sa.Text()), nullable=True + ), + sa.Column("imputation_method", sa.Text(), nullable=True), + sa.Column( + "ingested_at", + sa.DateTime(timezone=True), + server_default=sa.text("now()"), + nullable=False, + ), + sa.Column( + "raw_data", postgresql.JSONB(none_as_null=True, astext_type=sa.Text()), nullable=False + ), + sa.CheckConstraint( + "(source = 'csv' AND dataset_id IS NOT NULL) OR (source IN ('api_current', 'api_history') AND dataset_id IS NULL)", + name="ck_readings_dataset_source", + ), + sa.CheckConstraint( + "data_quality IS NULL OR data_quality IN ('good', 'partial', 'degraded', 'critical')", + name="ck_readings_quality", + ), + sa.CheckConstraint( + "source IN ('csv', 'api_current', 'api_history')", name="ck_readings_source" + ), + sa.CheckConstraint( + "(imputed_values IS NULL AND imputation_method IS NULL) OR (imputed_values IS NOT NULL AND imputation_method IS NOT NULL)", + name="ck_readings_imputation", + ), + sa.ForeignKeyConstraint( + ["dataset_id"], ["datasets.dataset_id"], name="fk_readings_dataset", ondelete="RESTRICT" + ), + sa.ForeignKeyConstraint( + ["site_id"], ["sites.site_id"], name="fk_readings_site", ondelete="RESTRICT" + ), + sa.PrimaryKeyConstraint("reading_id", "timestamp"), + ) + op.create_index("ix_readings_dataset_id", "readings", ["dataset_id"], unique=False) + op.create_index( + "ix_readings_site_timestamp", "readings", ["site_id", "timestamp"], unique=False + ) + op.create_index( + "uq_readings_source", + "readings", + ["site_id", "timestamp", "source", sa.literal_column("coalesce(dataset_id, 0)")], + unique=True, + ) + op.execute( + "SELECT create_hypertable('readings', by_range('timestamp'), create_default_indexes => FALSE)" + ) + op.create_table( + "alerts", + sa.Column("id", sa.BigInteger(), autoincrement=True, nullable=False), + sa.Column("alert_id", sa.Text(), nullable=False), + sa.Column("site_id", sa.Text(), nullable=False), + sa.Column("source", sa.Text(), nullable=False), + sa.Column("timestamp", sa.DateTime(timezone=True), nullable=False), + sa.Column("type", sa.Text(), nullable=False), + sa.Column("severity", sa.Text(), nullable=False), + sa.Column("message", sa.Text(), nullable=False), + sa.Column("value", sa.Double(), nullable=True), + sa.Column("threshold", sa.Double(), nullable=True), + sa.Column("metric", sa.Text(), nullable=True), + sa.Column("prediction_id", sa.BigInteger(), nullable=True), + sa.Column( + "raw_data", postgresql.JSONB(none_as_null=True, astext_type=sa.Text()), nullable=False + ), + sa.CheckConstraint( + "severity IN ('low', 'medium', 'high', 'critical')", name="ck_alerts_severity" + ), + sa.CheckConstraint("source IN ('api_mock', 'enervision')", name="ck_alerts_source"), + sa.CheckConstraint( + "type IN ('spike', 'threshold', 'anomaly', 'outage', 'sensor')", name="ck_alerts_type" + ), + sa.ForeignKeyConstraint( + ["prediction_id", "site_id"], + ["predictions.prediction_id", "predictions.site_id"], + name="fk_alerts_prediction_site", + ondelete="RESTRICT", + ), + sa.ForeignKeyConstraint( + ["site_id"], ["sites.site_id"], name="fk_alerts_site", ondelete="RESTRICT" + ), + sa.PrimaryKeyConstraint("id"), + sa.UniqueConstraint("source", "site_id", "alert_id", name="uq_alerts_source_site_id"), + ) + op.create_index("ix_alerts_site_timestamp", "alerts", ["site_id", "timestamp"], unique=False) + op.create_table( + "recommendations", + sa.Column("recommendation_id", sa.BigInteger(), autoincrement=True, nullable=False), + sa.Column("alert_id", sa.BigInteger(), nullable=False), + sa.Column("action", sa.Text(), nullable=False), + sa.Column("explanation", sa.Text(), nullable=False), + sa.Column("rule_reference", sa.Text(), nullable=False), + sa.Column( + "created_at", + sa.DateTime(timezone=True), + server_default=sa.text("now()"), + nullable=False, + ), + sa.ForeignKeyConstraint( + ["alert_id"], ["alerts.id"], name="fk_recommendations_alert", ondelete="RESTRICT" + ), + sa.PrimaryKeyConstraint("recommendation_id"), + sa.UniqueConstraint("alert_id", "rule_reference", name="uq_recommendations_alert_rule"), + ) + # ### end Alembic commands ### + + +def downgrade() -> None: + op.drop_table("recommendations") + op.drop_table("alerts") + op.drop_table("readings") + op.drop_table("predictions") + op.drop_table("sites") + op.drop_table("datasets") diff --git a/apps/backend/app/models/__init__.py b/apps/backend/app/models/__init__.py index 6d71227..0f48e79 100644 --- a/apps/backend/app/models/__init__.py +++ b/apps/backend/app/models/__init__.py @@ -1,2 +1,6 @@ # Piege : tout modele absent de ce module reste invisible de `alembic revision # --autogenerate`, qui genererait alors un drop de sa table. + +from app.models.energy import Alert, Dataset, Prediction, Reading, Recommendation, Site + +__all__ = ["Alert", "Dataset", "Prediction", "Reading", "Recommendation", "Site"] diff --git a/apps/backend/app/models/energy.py b/apps/backend/app/models/energy.py new file mode 100644 index 0000000..de27c7c --- /dev/null +++ b/apps/backend/app/models/energy.py @@ -0,0 +1,209 @@ +"""Tables du modèle de données EnerVision (CSV, API Mock et résultats ML).""" + +from datetime import datetime +from decimal import Decimal +from typing import Any + +from sqlalchemy import ( + BigInteger, + Boolean, + CheckConstraint, + DateTime, + Double, + ForeignKey, + ForeignKeyConstraint, + Index, + Integer, + Numeric, + String, + Text, + UniqueConstraint, + func, + text, +) +from sqlalchemy.dialects.postgresql import ARRAY, JSONB +from sqlalchemy.orm import Mapped, mapped_column + +from app.db.base import Base + + +class Dataset(Base): + __tablename__ = "datasets" + __table_args__ = ( + CheckConstraint("dataset_id > 0", name="ck_datasets_positive_id"), + UniqueConstraint("archive_sha256", name="uq_datasets_archive_sha256"), + ) + + dataset_id: Mapped[int] = mapped_column(BigInteger, primary_key=True, autoincrement=True) + dataset_name: Mapped[str] = mapped_column(Text) + archive_sha256: Mapped[str] = mapped_column(String(64)) + storage_uri: Mapped[str] = mapped_column(Text) + source_timezone: Mapped[str | None] = mapped_column(Text) + # "metadata" est réservé par SQLAlchemy ; le nom SQL reste inchangé. + dataset_metadata: Mapped[dict[str, Any]] = mapped_column("metadata", JSONB(none_as_null=True)) + + +class Site(Base): + __tablename__ = "sites" + + site_id: Mapped[str] = mapped_column(Text, primary_key=True) + site_name: Mapped[str] = mapped_column(Text) + site_type: Mapped[str] = mapped_column(Text) + location: Mapped[str | None] = mapped_column(Text) + capacity_kw: Mapped[float | None] = mapped_column(Double) + status: Mapped[str | None] = mapped_column(Text) + + +class Reading(Base): + __tablename__ = "readings" + __table_args__ = ( + CheckConstraint( + "source IN ('csv', 'api_current', 'api_history')", name="ck_readings_source" + ), + CheckConstraint( + "(source = 'csv' AND dataset_id IS NOT NULL) OR " + "(source IN ('api_current', 'api_history') AND dataset_id IS NULL)", + name="ck_readings_dataset_source", + ), + CheckConstraint( + "data_quality IS NULL OR data_quality IN ('good', 'partial', 'degraded', 'critical')", + name="ck_readings_quality", + ), + CheckConstraint( + "(imputed_values IS NULL AND imputation_method IS NULL) OR " + "(imputed_values IS NOT NULL AND imputation_method IS NOT NULL)", + name="ck_readings_imputation", + ), + Index("ix_readings_site_timestamp", "site_id", "timestamp"), + Index("ix_readings_dataset_id", "dataset_id"), + ) + + reading_id: Mapped[int] = mapped_column(BigInteger, primary_key=True, autoincrement=True) + site_id: Mapped[str] = mapped_column( + Text, ForeignKey("sites.site_id", name="fk_readings_site", ondelete="RESTRICT") + ) + timestamp: Mapped[datetime] = mapped_column(DateTime(timezone=True), primary_key=True) + source: Mapped[str] = mapped_column(Text) + dataset_id: Mapped[int | None] = mapped_column( + BigInteger, + ForeignKey("datasets.dataset_id", name="fk_readings_dataset", ondelete="RESTRICT"), + ) + consumption_kw: Mapped[float | None] = mapped_column(Double) + consumption_kwh: Mapped[float | None] = mapped_column(Double) + consumption_euros: Mapped[Decimal | None] = mapped_column(Numeric(14, 2)) + voltage_v: Mapped[float | None] = mapped_column(Double) + current_a: Mapped[float | None] = mapped_column(Double) + power_factor: Mapped[float | None] = mapped_column(Double) + temperature_celsius: Mapped[float | None] = mapped_column(Double) + humidity_percent: Mapped[float | None] = mapped_column(Double) + solar_irradiance_wm2: Mapped[float | None] = mapped_column(Double) + is_working_hours: Mapped[bool | None] = mapped_column(Boolean) + data_quality: Mapped[str | None] = mapped_column(Text) + null_reasons: Mapped[list[str] | None] = mapped_column(ARRAY(Text)) + imputed_values: Mapped[dict[str, Any] | None] = mapped_column(JSONB(none_as_null=True)) + imputation_method: Mapped[str | None] = mapped_column(Text) + ingested_at: Mapped[datetime] = mapped_column( + DateTime(timezone=True), server_default=func.now() + ) + raw_data: Mapped[dict[str, Any]] = mapped_column(JSONB(none_as_null=True)) + + +Index( + "uq_readings_source", + Reading.site_id, + Reading.timestamp, + Reading.source, + func.coalesce(Reading.dataset_id, text("0")), + unique=True, +) + + +class Prediction(Base): + __tablename__ = "predictions" + __table_args__ = ( + UniqueConstraint("prediction_id", "site_id", name="uq_predictions_id_site"), + Index("ix_predictions_site_target", "site_id", "target_at"), + CheckConstraint( + "target_metric IN ('consumption_kwh', 'consumption_kw')", + name="ck_predictions_metric", + ), + CheckConstraint( + "period_minutes IS NULL OR period_minutes > 0", name="ck_predictions_period" + ), + CheckConstraint( + "target_metric <> 'consumption_kwh' OR period_minutes IS NOT NULL", + name="ck_predictions_energy_period", + ), + CheckConstraint( + "(status = 'available' AND predicted_value IS NOT NULL AND failure_reason IS NULL) OR " + "(status IN ('insufficient_data', 'error') AND predicted_value IS NULL " + "AND failure_reason IS NOT NULL)", + name="ck_predictions_status", + ), + ) + + prediction_id: Mapped[int] = mapped_column(BigInteger, primary_key=True, autoincrement=True) + site_id: Mapped[str] = mapped_column( + Text, ForeignKey("sites.site_id", name="fk_predictions_site", ondelete="RESTRICT") + ) + created_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), server_default=func.now()) + target_at: Mapped[datetime] = mapped_column(DateTime(timezone=True)) + target_metric: Mapped[str] = mapped_column(Text) + period_minutes: Mapped[int | None] = mapped_column(Integer) + predicted_value: Mapped[float | None] = mapped_column(Double) + model_reference: Mapped[str] = mapped_column(Text) + status: Mapped[str] = mapped_column(Text) + failure_reason: Mapped[str | None] = mapped_column(Text) + + +class Alert(Base): + __tablename__ = "alerts" + __table_args__ = ( + UniqueConstraint("source", "site_id", "alert_id", name="uq_alerts_source_site_id"), + Index("ix_alerts_site_timestamp", "site_id", "timestamp"), + ForeignKeyConstraint( + ["prediction_id", "site_id"], + ["predictions.prediction_id", "predictions.site_id"], + name="fk_alerts_prediction_site", + ondelete="RESTRICT", + ), + CheckConstraint("source IN ('api_mock', 'enervision')", name="ck_alerts_source"), + CheckConstraint( + "type IN ('spike', 'threshold', 'anomaly', 'outage', 'sensor')", name="ck_alerts_type" + ), + CheckConstraint( + "severity IN ('low', 'medium', 'high', 'critical')", name="ck_alerts_severity" + ), + ) + + id: Mapped[int] = mapped_column(BigInteger, primary_key=True, autoincrement=True) + alert_id: Mapped[str] = mapped_column(Text) + site_id: Mapped[str] = mapped_column( + Text, ForeignKey("sites.site_id", name="fk_alerts_site", ondelete="RESTRICT") + ) + source: Mapped[str] = mapped_column(Text) + timestamp: Mapped[datetime] = mapped_column(DateTime(timezone=True)) + type: Mapped[str] = mapped_column(Text) + severity: Mapped[str] = mapped_column(Text) + message: Mapped[str] = mapped_column(Text) + value: Mapped[float | None] = mapped_column(Double) + threshold: Mapped[float | None] = mapped_column(Double) + metric: Mapped[str | None] = mapped_column(Text) + prediction_id: Mapped[int | None] = mapped_column(BigInteger) + raw_data: Mapped[dict[str, Any]] = mapped_column(JSONB(none_as_null=True)) + + +class Recommendation(Base): + __tablename__ = "recommendations" + __table_args__ = ( + UniqueConstraint("alert_id", "rule_reference", name="uq_recommendations_alert_rule"), + ) + + recommendation_id: Mapped[int] = mapped_column(BigInteger, primary_key=True, autoincrement=True) + alert_id: Mapped[int] = mapped_column( + BigInteger, ForeignKey("alerts.id", name="fk_recommendations_alert", ondelete="RESTRICT") + ) + action: Mapped[str] = mapped_column(Text) + explanation: Mapped[str] = mapped_column(Text) + rule_reference: Mapped[str] = mapped_column(Text) + created_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), server_default=func.now())