---

Inventory Optimization Tool

This project provides an inventory optimization tool that uses synthetic data and applies demand forecasting, inventory classification, safety stock calculation, and other inventory management techniques. It is containerized using Docker to ensure easy deployment and portability.

Blog

a detailed write up to understand the tool in-depth.

Requirements

Before you start, ensure you have the following installed:

• Docker: Get Docker

How to Run

Step 1: Clone the Repository

Clone the project to your local machine.

git clone https://github.com/saru2020/InventoryOptimizer.git
cd InventoryOptimizer

Step 2: Add Data Directory

Make sure to have a /data folder at the root of the project to store input and output CSV files.

mkdir data

Step 3: Build and Run the Docker Container

Use the provided run.sh script to build and run the Docker container. This will automatically build the image, start the container, and map the /data folder for storing input and output files.

chmod +x run.sh
./run.sh

Alternatively, you can manually run the commands:

# Build Docker image
docker build -t inventory_optimizer .

# Run Docker container
docker run -v $(pwd)/data:/data --name inventory_optimizer inventory_optimizer

The container will generate synthetic data and save the results to the /data/output.csv and /data/store_turnover.csv files.

Step 4: View Results

Once the container has finished processing, the results will be stored in the /data folder. The main outputs are:

• output.csv: Contains the daily inventory orders, forecast data, and other information for each SKU and store.
• store_turnover.csv: Contains inventory turnover ratios for each store.

Comprehensive Inventory Metrics

This enhanced version now supports ALL 16 comprehensive inventory metrics:

📖 For detailed explanations, formulas, and implementation details of all metrics, see METRICS_DOCUMENTATION.md

Core Inventory Optimization Metrics (6)

1. Safety Stock 🛡️ - Buffer inventory to prevent stock-outs
2. Economic Order Quantity (EOQ) 📦 - Optimal order quantity for cost optimization
3. Reorder Point 🔄 - Inventory level for timely reordering (with pending orders)
4. Daily Order Quantity 📋 - Daily ordering decisions with priority ranking
5. Inventory Turnover 🔄 - Efficiency measurement per SKU
6. Store Turnover 🏪 - Store-level efficiency comparison

Advanced Analytics Metrics (10)

7. Excess Inventory Items 📦 - Overstocked items identification
8. Stock-out Risk Items ⚠️ - Items at risk of running out
9. Never Sold Items 🚫 - Dead stock identification
10. Fast/Slow Moving Items 🏃‍♂️ - Velocity-based classification
11. Most Urgent Orders Items 🚨 - Priority ranking for reorders
12. Most Profitable/Stagnant Items 💰 - Value-based focus
13. Days Left for Stock-out ⏰ - Time-based urgency
14. Average Daily Demand 📊 - Demand baseline for forecasting
15. Lead Time Analysis ⏱️ - Supplier performance tracking
16. On Order Analysis 📋 - Pending order tracking

Metrics Summary

#	Metric	Category	Formula	Business Impact
1	Safety Stock	Core	Z-score × √(Lead Time) × Std Dev	Prevents stock-outs
2	EOQ	Core	√(2 × Demand × Order Cost / Holding Cost)	Cost optimization
3	Reorder Point	Core	Safety Stock + (Demand × Lead Time) - Pending	Timely reordering
4	Daily Order	Core	max(Reorder Point - Inventory - Pending, 0)	Daily decisions
5	Inventory Turnover	Core	Total Sales / Average Inventory	Efficiency measure
6	Store Turnover	Core	Total Sales / Average Inventory Value	Store comparison
7	Excess Inventory	Analytics	Current - (Safety + Demand × Lead Time)	Capital optimization
8	Stock-out Risk	Analytics	(Safety + Demand × Lead Time) - Current	Prevents lost sales
9	Never Sold	Analytics	Total Sales = 0	Dead stock identification
10	Fast/Slow Moving	Analytics	Inventory Turnover classification	Velocity optimization
11	Urgent Orders	Analytics	(Reorder Point - Current) / Demand	Priority ranking
12	Profitable/Stagnant	Analytics	Profit per Unit × Total Sales	Value focus
13	Days to Stock-out	Analytics	Current Inventory / Daily Demand	Time urgency
14	Average Daily Demand	Analytics	Total Sales / Period Days	Demand baseline
15	Lead Time Analysis	Analytics	Order placement to receipt time	Supplier performance
16	On Order Analysis	Analytics	Sum of pending order quantities	Order tracking

Example Data

Enhanced Input Schema

The enhanced script generates comprehensive synthetic data with all required fields:

Date	Store	SKU	SalesQuantity	InventoryLevel	UnitCost	SellingPrice	LeadTime	SupplierID	Category	OrderedDate	OrderedQuantity	OrderStatus
2023-01-01	Store001	SKU001	10	55	45.50	89.99	7	SUP001	Electronics	2022-12-25	50	Shipped
2023-01-01	Store001	SKU002	14	42	25.00	49.99	5	SUP002	Clothing	2022-12-28	30	Confirmed
2023-01-01	Store001	SKU003	9	75	15.00	29.99	3	SUP003	Books	null	0	None

Enhanced Output Schema

After processing, the tool outputs comprehensive inventory data with all metrics:

Date	Store	SKU	OrderQuantity	InventoryTurnover	DemandForecast	SafetyStock	ReorderPoint	Reorder	PendingOrders	OrderPriority	ExcessInventory	StockoutRisk	DaysToStockout
2023-01-01	Store001	SKU001	8	2.5	10.2	8.5	28.5	True	50	High	-23.5	-23.5	5.4
2023-01-01	Store001	SKU002	15	1.8	14.1	9.2	39.7	False	30	Medium	-37.7	37.7	3.0
2023-01-01	Store001	SKU003	0	0.9	8.7	7.8	32.1	False	0	Low	42.9	32.1	8.6

Store Turnover Output

Store	InventoryTurnover
Store001	2.0
Store002	1.7
Store003	1.5

Comprehensive Metrics Output

The tool now generates 13 separate CSV files for detailed analysis:

Core Output Files:

1. output.csv - Main output with all calculated metrics
2. source.csv - Enhanced input data with all required fields
3. store_turnover.csv - Store-level inventory turnover ratios

Advanced Analytics Files: 4. excess_inventory_analysis.csv - Excess inventory analysis 5. stockout_risk_analysis.csv - Stock-out risk analysis
6. never_sold_analysis.csv - Never sold items analysis 7. fast_slow_moving_analysis.csv - Movement classification 8. urgent_orders_analysis.csv - Urgent orders analysis 9. profitable_stagnant_analysis.csv - Profitability analysis 10. days_to_stockout_analysis.csv - Days to stock-out analysis 11. average_daily_demand_analysis.csv - Demand analysis 12. lead_time_analysis_analysis.csv - Lead time analysis 13. on_order_analysis_analysis.csv - On order analysis

Configuration Options

Holiday Configuration

The system now supports configurable holidays by country and state:

# Example usage with different countries/states
main(country='US', state='CA', start_date='2023-01-01', end_date='2023-12-31')
main(country='GB', state='', start_date='2023-01-01', end_date='2023-12-31')
main(country='IN', state='TN', start_date='2023-01-01', end_date='2023-12-31')

Data Schema

The enhanced input schema includes all required fields for comprehensive analysis:

Required Columns:

• Date: Transaction date
• Store: Store identifier
• SKU: Product identifier
• SalesQuantity: Quantity sold
• InventoryLevel: Current inventory level
• UnitCost: Cost per unit
• SellingPrice: Selling price per unit
• LeadTime: Lead time in days
• SupplierID: Supplier identifier
• Category: Product category
• OrderedDate: Date when order was placed (for lead time calculation)
• OrderedQuantity: Quantity ordered (for reorder point calculation)
• OrderStatus: Status of the order (Pending, Confirmed, Shipped, In Transit, Delivered, None)

Testing and Quality Assurance

The project includes comprehensive testing:

# Run all tests
python test_inventory_optimization.py

# Setup pre-commit hooks (runs tests automatically before commits)
python setup_precommit.py

# Run pre-commit on all files
pre-commit run --all-files

File Outputs

• source.csv: Enhanced input data with all required fields
• output.csv: Main output with all calculated metrics
• store_turnover.csv: Store-level inventory turnover ratios
• *_analysis.csv: 10 separate files for detailed metric analysis
• METRICS_DOCUMENTATION.md: Detailed explanations of all 16 metrics

Total: 13 CSV files + 1 documentation file providing comprehensive inventory insights

Advanced Usage

Custom Data Input

To use your own data, ensure your CSV has the required schema and modify the data loading:

# Load your own data
def load_custom_data(file_path):
    df = pd.read_csv(file_path)
    # Ensure all required columns exist
    required_columns = ['Date', 'Store', 'SKU', 'SalesQuantity', 'InventoryLevel', 
                       'UnitCost', 'SellingPrice', 'LeadTime', 'SupplierID', 'Category', 'OrderedDate']
    for col in required_columns:
        if col not in df.columns:
            raise ValueError(f"Missing required column: {col}")
    return df

Metric Customization

All metrics support customization:

# Custom excess inventory threshold
excess_df = calculate_excess_inventory_items(df, excess_threshold=2.0)

# Custom movement classification thresholds
movement_df = calculate_fast_slow_moving_items(df)
# Modify thresholds in the function as needed

---

Feel free to adapt this to include more examples or additional details based on your specific project requirements!