Initial commit

omop/scripts/load_sample_data.sh

@@ -0,0 +1,80 @@
+#!/bin/bash
+# Load Sample Data Script
+# This script sets up the database and loads sample data for testing
+
+set -e
+
+echo "=========================================="
+echo "OMOP Sample Data Loading Script"
+echo "=========================================="
+echo ""
+
+# Colors for output
+GREEN='\033[0;32m'
+YELLOW='\033[1;33m'
+RED='\033[0;31m'
+NC='\033[0m' # No Color
+
+# Check if we're in the right directory
+if [ ! -f "setup.py" ]; then
+    echo -e "${RED}Error: Must be run from omop directory${NC}"
+    exit 1
+fi
+
+# Step 1: Install dependencies
+echo -e "${YELLOW}Step 1: Installing dependencies...${NC}"
+pip install faker > /dev/null 2>&1 || echo "Faker already installed"
+echo -e "${GREEN}✓ Dependencies installed${NC}"
+echo ""
+
+# Step 2: Create database schemas
+echo -e "${YELLOW}Step 2: Creating database schemas...${NC}"
+python -m src.cli.commands schema create --type all 2>/dev/null || echo "Schemas may already exist"
+echo -e "${GREEN}✓ Schemas ready${NC}"
+echo ""
+
+# Step 3: Generate and load sample data
+echo -e "${YELLOW}Step 3: Generating and loading sample data...${NC}"
+python scripts/generate_sample_data.py
+echo -e "${GREEN}✓ Sample data loaded${NC}"
+echo ""
+
+# Step 4: Verify data
+echo -e "${YELLOW}Step 4: Verifying loaded data...${NC}"
+python -c "
+from src.utils.config import Config
+from src.utils.db_connection import DatabaseConnection
+from sqlalchemy import text
+
+config = Config.load('config.yaml')
+db = DatabaseConnection(config)
+
+with db.get_session() as session:
+    # Count records in staging tables
+    tables = ['raw_patients', 'raw_visits', 'raw_conditions', 'raw_drugs']
+    
+    print('\nStaging Table Counts:')
+    print('-' * 40)
+    for table in tables:
+        query = text(f'SELECT COUNT(*) FROM staging.{table}')
+        count = session.execute(query).fetchone()[0]
+        print(f'  staging.{table:20s}: {count:5d} records')
+    print('-' * 40)
+"
+echo -e "${GREEN}✓ Data verification complete${NC}"
+echo ""
+
+echo "=========================================="
+echo -e "${GREEN}Sample data loading complete!${NC}"
+echo "=========================================="
+echo ""
+echo "Next steps:"
+echo "  1. Run ETL pipeline:"
+echo "     omop-pipeline etl run --source staging.raw_patients --target person"
+echo ""
+echo "  2. View statistics:"
+echo "     omop-pipeline stats show"
+echo ""
+echo "  3. Validate data:"
+echo "     omop-pipeline validate"
+echo ""