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

# Business Intelligence Dashboard

An interactive Business Intelligence dashboard built with Gradio that enables users to explore and analyze business data through an intuitive, Tableau-like web interface.

## Features

### 📁 Data Upload & Validation
- Upload CSV or Excel files through the web interface
- Display basic dataset information (shape, columns, data types)
- Show data preview (first 10 rows)
- Graceful error handling with informative messages

### 📈 Data Exploration & Summary Statistics
- **Automated Data Profiling:**
  - Numerical columns: mean, median, std, min, max, quartiles
  - Categorical columns: unique values, value counts, mode
  - Missing value report
  - Correlation matrix for numerical features

### 🔍 Interactive Filtering
- Dynamic filtering interface based on column types:
  - **Numerical:** Range sliders with min/max inputs
  - **Categorical:** Multi-select checkboxes
  - **Date:** Date range pickers (when applicable)
- Real-time row count updates as filters are applied
- Display filtered data preview

### 📊 Visualizations
Implements 5 different visualization types:
1. **Time Series Plot:** Trends over time with aggregation options
2. **Distribution Plot:** Histogram or box plot for numerical data
3. **Category Analysis:** Bar chart or pie chart for categorical data
4. **Scatter Plot:** Show relationships between variables
5. **Correlation Heatmap:** Visualize correlations between numerical features

**Features:**
- User selects which columns to visualize
- Clear titles, labels, and legends
- Multiple aggregation methods (sum, mean, count, median)
- Professional Plotly visualizations

### 💡 Insights Generation
Automatically generates insights:
- **Top/Bottom Performers:** Identify highest/lowest values
- **Basic Trends:** Detect patterns in time series data
- **Summary Statistics:** High-level dataset overview

### 💾 Export Functionality
- Export filtered data as CSV
- Export visualizations as PNG images

## High-Level Architecture

```
┌─────────────────────────────────────────────────────────────────┐
│                         User Interface                          │
│                      (Gradio Web Interface)                     │
│  ┌──────────────┐  ┌──────────────┐  ┌──────────────┐           │
│  │ Data Upload  │  │ Visualization│  │   Insights   │           │
│  │   & Preview  │  │   & Charts   │  │  Generation  │           │
│  └──────────────┘  └──────────────┘  └──────────────┘           │
│  ┌──────────────┐  ┌──────────────┐  ┌──────────────┐           │
│  │  Statistics  │  │   Filter &   │  │    Export    │           │
│  │  & Profiling │  │   Explore    │  │ Functionality│           │
│  └──────────────┘  └──────────────┘  └──────────────┘           │
└────────────────────────────┬────────────────────────────────────┘
                             │
                             ▼
┌─────────────────────────────────────────────────────────────────┐
│                    Application Layer (app.py)                   │
│  • Orchestrates user interactions                               │
│  • Manages global state (current_df, filters, figures)          │
│  • Routes requests to appropriate modules                       │
└────────────────────────────┬────────────────────────────────────┘
                             │
        ┌────────────────────┼────────────────────┐
        │                    │                    │
        ▼                    ▼                    ▼
┌──────────────────┐ ┌──────────────────┐ ┌──────────────────┐
│  Data Processing │ │  Visualizations  │ │     Insights     │
│   Layer          │ │     Layer        │ │      Layer       │
│                  │ │                  │ │                  │
│ data_processor.py│ │visualizations.py │ │  insights.py     │
│                  │ │                  │ │                  │
│ • CSV/Excel Load │ │ • Time Series    │ │ • Top/Bottom     │
│ • Data Cleaning  │ │ • Distribution   │ │   Performers     │
│ • Filtering      │ │ • Category       │ │ • Trend Analysis │
│ • Statistics     │ │   Analysis       │ │ • Summary Stats  │
│   Generation     │ │ • Scatter Plot   │ │                  │
│                  │ │ • Correlation    │ │                  │
│                  │ │   Heatmap        │ │                  │
└──────────────────┘ └──────────────────┘ └──────────────────┘
        │                    │                    │
        └────────────────────┼────────────────────┘
                             │
                             ▼
┌─────────────────────────────────────────────────────────────────┐
│                    Utilities Layer (utils.py)                   │
│  • Column type detection (numerical, categorical, date)         │
│  • Missing value analysis                                       │
│  • Data validation helpers                                      │
└────────────────────────────┬────────────────────────────────────┘
                             │
                             ▼
┌─────────────────────────────────────────────────────────────────┐
│                      Data Sources                               │
│  ┌──────────────┐  ┌──────────────┐  ┌──────────────┐           │
│  │ stocks.csv   │  │sales_train   │  │Online Retail │           │
│  │              │  │   .csv       │  │   .xlsx      │           │
│  └──────────────┘  └──────────────┘  └──────────────┘           │
│                                                                 │
│  • CSV files (pandas.read_csv)                                  │
│  • Excel files (pandas.read_excel)                              │
│  • User-uploaded datasets                                       │
└─────────────────────────────────────────────────────────────────┘

┌─────────────────────────────────────────────────────────────────┐
│                    External Libraries                           │
│  • pandas: Data manipulation and analysis                       │
│  • plotly: Interactive visualizations                           │
│  • gradio: Web interface framework                              │
│  • numpy: Numerical computations                                │
└─────────────────────────────────────────────────────────────────┘
```

## Project Structure

```
project/
├── 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 
├── constants.py           # Constants used throughout the code
├── requirements.txt       # Python dependencies
├── README.md              # This file
└── data/                  # Sample datasets
    ├── sales_train.csv
    ├── stocks.csv
    └── Online Retail.xlsx
```

## Setup Instructions

### 1. Install Dependencies

```bash
pip install -r requirements.txt
```

**Note:** This project uses Gradio 6.0.2, which includes improved performance and updated APIs. Make sure you have Python 3.8 or higher installed.

### 2. Run the Application

```bash
python app.py
```

The application will launch and be accessible at `http://localhost:7860` in your web browser.

## Usage

1. **Upload Data:** Navigate to the "Data Upload & Preview" tab and upload a CSV or Excel file
2. **View Statistics:** Go to "Statistics & Profiling" to see comprehensive data statistics
3. **Apply Filters:** Use "Filter & Explore" to filter your data by column values
4. **Create Visualizations:** Visit "Visualizations" to create interactive charts
5. **Generate Insights:** Check "Insights" for automated data insights
6. **Export Data:** Use "Export" to download filtered data or visualizations

## Aggregation Methods

The dashboard supports multiple aggregation methods for visualizations:
- **Sum**: Adds all values together (useful for totals, volumes)
- **Mean**: Calculates the average value (useful for prices, rates)
- **Count**: Counts the number of data points (useful for frequency)
- **Median**: Finds the middle value (robust to outliers)
- **None**: No aggregation (shows raw data points)

## Step-by-Step Tutorial: Monthly Average Closing Price

Let's walk through a complete example:

### Step 1: Load the Data
1. Open the dashboard
2. Go to **📁 Data Upload & Preview** tab
3. Click **Upload Dataset**
4. Select `sample-datasets/stocks.csv`
5. Click **Load Data**
6. Verify the data preview shows the stock data

### Step 2: Create the Visualization
1. Navigate to **📊 Visualizations** tab
2. Configure the chart:
   - **Chart Type**: `Time Series`
   - **X-Axis Column**: `Date`
   - **Y-Axis Column**: `Close`
   - **Aggregation Method**: `Mean`
3. Click **Generate Visualization**

### Step 3: Interpret the Results
- The chart shows a line graph with dates on X-axis and average closing prices on Y-axis
- Each point represents the mean closing price for that date
- You can see trends, patterns, and changes over time

### Step 4: Compare Different Aggregations
Try generating the same chart with different aggregation methods:
- **Mean**: Average closing price (smooth trend)
- **Sum**: Total closing price (not meaningful for prices, but shows concept)
- **Median**: Middle closing price (robust to outliers)
- **None**: All individual closing prices (may be cluttered)

## Technical Details

### Design Patterns

The application uses the **Strategy Pattern** for:
- **Data Loading:** Different strategies for CSV vs Excel files
- **Data Filtering:** Different strategies for numerical, categorical, and date filters
- **Visualizations:** Different strategies for each chart type

### Code Quality

- Follows PEP 8 style guidelines
- Comprehensive docstrings for all functions
- Proper error handling with try/except blocks
- Modular design with clear separation of concerns
- No hardcoded values (uses constants and configuration)

### Libraries

- **pandas 2.2.0+:** All data manipulation and analysis
- **Gradio 6.0.2:** Web interface framework
- **Plotly 5.22.0+:** Interactive visualizations
- **matplotlib 3.8.0+ / seaborn 0.13.0+:** Additional visualization support
- **Python 3.8+:** Following best practices

## Sample Datasets

The `data/` folder includes sample datasets:
- `sales_train.csv`: Sales transaction data
- `stocks.csv`: Stock market data
- `Online Retail.xlsx`: E-commerce retail data

## Requirements

- Python 3.8 or higher
- All dependencies listed in `requirements.txt`:
  - pandas >= 2.2.0
  - numpy >= 1.26.0
  - gradio == 6.0.2
  - matplotlib >= 3.8.0
  - seaborn >= 0.13.0
  - plotly >= 5.22.0
  - kaleido >= 0.2.1
  - openpyxl >= 3.1.5
  - Pillow >= 10.4.0

## License

This project is created for educational purposes as part of CS5130 coursework.