🧾 Synthetic Data Generator — Phase 1 (Single Table Edition)

Welcome to the Synthetic Data Generator — a modular, production-ready Python platform to generate realistic synthetic datasets for analytics, testing, and ML workflows.

---

🚀 Features (Phase 1)

✅ Streamlit-Based Interactive UI

• Fully editable column builder in spreadsheet-style format
• Supports dynamic column count & custom names
• Live schema preview (JSON view)

✅ Supported Column Types

Data Type	Input Format Example	Notes
Integer	0, 100	Range-based values
Float	10000.0, 200000.0	Range-based float values
Categorical	Data, Dev, PM, Support	Comma-separated list of categories
PII	name, email, phone, address	Uses faker to generate synthetic PII
Date	01-01-2020, 12-31-2024	Format: MM-DD-YYYY (auto ISO handled)
Conditional	Python logic per row (see below)	E.g. if salary > 100000: grade = 'A'

---

🧠 Conditional Logic Support

Define column values using Python-style expressions:

✅ Single-line:

np.random.choice(['yes', 'no'])

✅ Multi-line logic:

if department == 'Data':
    salary = np.random.randint(40000, 120000)
elif department == 'PM':
    salary = np.random.randint(50000, 150000)

You may reference other columns using:

• row['department']
• df['name'] (for sampling)
• np.random.* functions (safe subset)

---

🧱 Folder Layout

For detailed breakdown see: architecture_overview.md

synthetic_data_generator/
├── app.py
├── config/         # schema loading/saving
├── generator/      # core generation logic
├── ui/             # Streamlit UI components
├── utils/          # helpers, validation, faker wrappers
├── assets/         # sample schemas
├── export/         # exported files
├── requirements.txt
└── README.md

---

🔧 Setup Instructions

# 1. Create virtual environment
python -m venv venv
source venv/bin/activate  # or venv\Scripts\activate (Windows)

# 2. Install dependencies
pip install -r requirements.txt

# 3. Run Streamlit app
streamlit run app.py

---

📤 Output & Export

• First 10 rows previewed in UI
• Export to:
  • CSV
  • JSON
  • Excel (.xlsx)
• Download buttons available directly in UI

---

📦 Schema Save/Load

• Save current schema as .json
• Re-upload anytime to resume editing
• Samples in assets/sample_schemas/

---

🛣️ Phase 2 Sneak Peek

• Multi-table SQL-style schema generation
• Define foreign key relationships
• Export as SQL dump or multi-table ZIP
• Cross-table logic, referential integrity

---

🧠 Who is this for?

• Data Scientists and ML Engineers
• QA and Testing Teams
• Data Engineers / ETL Mocks
• Business Analysts building dashboards

---

🔖 Tag: v1.0.0 — Last updated: Phase 1 Final Patch (Release date: 2025-06-26)