README.md

---
title: Business Intelligence Dashboard
emoji: 📊
colorFrom: blue
colorTo: indigo
sdk: gradio
sdk_version: 6.0.2
app_file: app.py
pinned: false
license: mit
---

# Business Intelligence Dashboard

A professional, interactive Business Intelligence dashboard built with Gradio that enables non-technical stakeholders to explore and analyze business data through an intuitive web interface.

## 🌟 Features

### Data Upload & Validation
- Support for CSV and Excel files (.xlsx, .xls)
- Automatic data type detection
- Data preview and basic information display
- Comprehensive error handling

### Data Exploration & Summary Statistics
- Automated data profiling for numerical and categorical columns
- Missing value analysis
- Correlation matrix for numerical features
- Descriptive statistics (mean, median, std, quartiles)

### Interactive Filtering
- **Numerical filters**: Range sliders with min/max inputs
- **Categorical filters**: Multi-select dropdowns
- **Date filters**: Date range selection
- Real-time row count updates
- Filtered data preview

### Visualizations (7 Types)
1. **Time Series Plot**: Trends over time with multiple aggregation methods
2. **Distribution (Histogram)**: Data distribution with mean/median lines
3. **Distribution (Box Plot)**: Statistical distribution with quartiles
4. **Bar Chart**: Category analysis with top N selection
5. **Pie Chart**: Proportional category breakdown
6. **Scatter Plot**: Relationship between variables with trend lines
7. **Correlation Heatmap**: Numerical feature correlations

All visualizations support:
- User-selectable columns
- Multiple aggregation methods (sum, mean, count, median)
- Professional styling with clear labels and legends

### Automated Insights
- **Top/Bottom Performers**: Identify highest/lowest values
- **Trends & Patterns**: Detect increasing/decreasing trends
- **Anomaly Detection**: Find outliers using statistical methods
- **Data Quality Assessment**: Completeness, uniqueness, consistency checks

### Export Functionality
- Export filtered data as CSV
- Export visualizations as high-resolution PNG images

## 📋 Requirements

```
gradio>=6.0.2
pandas>=2.0.0
matplotlib>=3.7.0
seaborn>=0.12.0
plotly>=5.14.0
openpyxl>=3.1.0
numpy>=1.24.0
scipy>=1.10.0
```

## 🚀 Installation

1. **Clone or download this repository**

2. **Install dependencies**:
```bash
pip install -r requirements.txt
```

## 💻 Usage

### Running the Application

```bash
python app.py
```

The dashboard will launch at `http://127.0.0.1:7860`

### Using the Dashboard

1. **Upload Data**: 
   - Navigate to the "Data Upload" tab
   - Click "Upload File" and select a CSV or Excel file
   - Click "Load Data" to process the file

2. **View Statistics**:
   - Go to the "Statistics" tab
   - Click "Generate Statistics" to see comprehensive data analysis

3. **Filter Data**:
   - Use the "Filter & Explore" tab
   - Select columns and apply filters
   - View filtered data preview and row counts

4. **Create Visualizations**:
   - Navigate to "Visualizations" tab
   - Select visualization type
   - Choose columns and aggregation methods
   - Click "Create Visualization"

5. **Generate Insights**:
   - Go to "Insights" tab
   - Click "Generate Insights" for automated analysis

6. **Export Results**:
   - Use the "Export" tab to download filtered data
   - Export visualizations from the Visualizations tab

## 📁 Project Structure

```
APDP_DASHBOARD/
│
├── app.py                  # Main Gradio application
├── data_processor.py       # Data loading, cleaning, filtering
├── visualizations.py       # Chart creation functions
├── insights.py            # Automated insight generation
├── utils.py               # Helper functions
├── requirements.txt        # Dependencies
├── README.md              # This file
└── data/                  # Sample datasets
    ├── sample_sales_data.csv
    └── sample_employee_data.csv
```

## 📊 Sample Datasets

### 1. Sales Data (`sample_sales_data.csv`)
- **Columns**: Date, Product, Category, Region, Sales, Units, Customer_Satisfaction
- **Rows**: 75 records
- **Use Case**: Sales analysis, time series trends, regional performance

### 2. Employee Data (`sample_employee_data.csv`)
- **Columns**: Employee_ID, Name, Department, Position, Salary, Years_Experience, Performance_Score, Projects_Completed, Training_Hours
- **Rows**: 50 records
- **Use Case**: HR analytics, salary analysis, performance evaluation

## 🔧 Code Architecture

![System Architecture](assets/architecture.svg)

### Modular Design

- **`utils.py`**: Utility functions for data formatting, validation, and type detection
- **`data_processor.py`**: Core data operations (loading, cleaning, filtering, aggregation)
- **`visualizations.py`**: Interactive Plotly-based chart creation with hover details
- **`insights.py`**: Advanced AI-driven insights, Smart Dashboard logic, and Comparison tools
- **`app.py`**: Gradio interface orchestrating all components

### Key Design Principles

1. **Separation of Concerns**: Each module has a specific responsibility
2. **Error Handling**: Comprehensive try-except blocks with user-friendly messages
3. **Type Hints**: All functions include type annotations
4. **Documentation**: Detailed docstrings for all functions
5. **PEP 8 Compliance**: Follows Python style guidelines

## 🎯 Features Checklist

- ✅ Data Upload & Validation (CSV/Excel support)
- ✅ Data Preview (first/last N rows)
- ✅ Summary Statistics (numerical & categorical)
- ✅ Missing Value Analysis
- ✅ Correlation Matrix
- ✅ Interactive Filtering (numerical, categorical, datetime)
- ✅ 7 Visualization Types
- ✅ Aggregation Methods (sum, mean, count, median)
- ✅ Automated Insights Generation
- ✅ Top/Bottom Performers
- ✅ Trend Detection
- ✅ Anomaly Detection
- ✅ Export Filtered Data (CSV)
- ✅ Export Visualizations (PNG)

## 🐛 Error Handling

The application handles:
- Invalid file formats
- Missing or corrupted data
- Empty datasets
- Invalid column selections
- Type mismatches
- Missing values in calculations

All errors display user-friendly messages in the interface.

## 🔍 Technical Highlights

### Data Processing
- Automatic datetime detection and conversion
- Flexible filtering system supporting multiple data types
- Memory-efficient operations for large datasets

### Visualizations
- Interactive Plotly charts with hover tooltips
- Automatic color schemes and responsive layout
- Statistical annotations (mean, median, trend lines)
- Support for grouped and aggregated data

### Insights Engine
- Z-score based anomaly detection
- Time series trend analysis
- Categorical pattern recognition
- Data quality scoring

## 📝 Best Practices

1. **Data Quality**: Clean your data before uploading for best results
2. **File Size**: For large files (>100MB), consider filtering in Excel first
3. **Datetime Columns**: Use standard formats (YYYY-MM-DD) for automatic detection
4. **Missing Values**: The app handles them gracefully, but review the missing value report

## 🤝 Contributing

This project follows PEP 8 style guidelines. When contributing:
- Add docstrings to all functions
- Include type hints
- Handle errors gracefully
- Write modular, reusable code

## 📄 License

This project is created for educational purposes as part of a data science course.

## 🙋 Support

For issues or questions:
1. Check the error messages in the interface
2. Review the sample datasets for format examples
3. Ensure all dependencies are installed correctly

## 🎓 Learning Outcomes

This project demonstrates:
- Data manipulation with pandas
- Interactive web applications with Gradio
- Data visualization best practices
- Statistical analysis and insight generation
- Clean, modular Python code architecture
- Error handling and user experience design

---

**Built with ❤️ using Python, Gradio, and pandas**