DataMimic.io: Advanced Synthetic Data Generation & No-Code EDA/Pre-processing Platform

Project Goal

To develop a comprehensive web application, DataMimic.io, that serves as a powerful, user-friendly tool for both generating highly customizable synthetic datasets and performing intuitive, no-code Exploratory Data Analysis (EDA) and data pre-processing on uploaded files. The platform will empower data scientists, analysts, students, and learners by simplifying complex data tasks.

Technologies Used

• Backend: Python (Flask, Pandas, NumPy, Scikit-learn, Faker)
• Frontend: HTML, CSS (Bootstrap 5), JavaScript

Features (Planned)

Module 1: Synthetic Data Generation

• Core Dataset Parameters: Number of Records, Missing Value Percentage, Feature Covariance/Correlation Control.
• Geographical Locality/Context: India, UK, US, Canada, Australia.
• Schema Definition & Customization:
  • Pre-defined Industry Schemas (Medical, Finance, Retail, Education, Automotive).
  • Custom Column Definition (various data types including Categorical).
• Download Options: CSV, Excel, JSON.

Module 2: No-Code EDA & Pre-processing

• File Upload Interface: Support for .csv and .xlsx files.
• Data Overview & Summary: General statistics, column-wise details (data type, missing values, unique values), numerical/categorical stats.
• (Highly Desirable Visualizations): Histograms, Bar charts, Correlation Heatmap.
• Data Pre-processing Operations:
  • Missing Value Handling (remove rows/cols, impute with mean/median/mode).
  • Duplicate Handling (remove duplicate rows).
  • Column Management (remove specific columns).
  • Data Type Conversion (Numeric, String, Date).
  • Data Scaling & Normalization (Min-Max, Standardization).
  • Text Cleaning (Inconsistent Capitalization Repair: UPPERCASE, lowercase, Title Case).

Setup and Running the Application

1. Clone the Repository (Once available)

git clone <your-repo-url>
cd DataMimic.io