data-science-mcp

jingchen0123/data-science-mcp

3.3

If you are the rightful owner of data-science-mcp and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to henry@mcphub.com.

A comprehensive Model Context Protocol (MCP) server for data science workflows, providing tools for data loading, exploration, visualization, processing, statistical analysis, and code generation.

The Data Science MCP Agent is a robust server designed to streamline data science workflows through the Model Context Protocol (MCP). It offers a suite of tools for data loading, exploration, visualization, processing, and statistical analysis, making it an essential resource for data scientists. The server supports CSV data management, allowing users to load, inspect, and save datasets efficiently. It provides comprehensive data quality assessments and exploratory data analysis, including missing value and outlier detection. Visualization capabilities include scatter plots, histograms, and correlation matrices, with customizable styling options. The server also facilitates complex data processing tasks such as filtering, transformation, and aggregation. Statistical analysis features include t-tests, ANOVA, and regression analysis, with power analysis and effect size calculations. Additionally, the server can generate Python code from natural language descriptions, execute it safely, and store code snippets for future use. Guided workflows and prompt templates assist users in common data science tasks, ensuring a smooth and efficient process.

Features

  • Data Exploration & Analysis: Load and inspect CSV datasets, perform data quality assessments, and conduct exploratory data analysis.
  • Data Visualization: Create scatter plots, histograms, bar charts, and line plots with customizable styling.
  • Data Processing: Filter datasets, transform columns, and handle missing values with various imputation methods.
  • Statistical Analysis: Perform correlation analysis, t-tests, ANOVA, and regression analysis.
  • Code Generation & Execution: Generate and execute Python analysis code from natural language descriptions.

Usages

usage with Claude Desktop

{
  "mcpServers": {
    "data-science": {
      "command": "python",
      "args": ["/absolute/path/to/data-science-mcp/server.py"],
      "env": {}
    }
  }
}

Tools

  1. load_csv

    Load CSV files into the system

  2. explore_data

    Comprehensive data exploration

  3. plot_scatter

    Create scatter plots

  4. run_ttest

    Perform one-sample t-test

  5. generate_analysis_code

    Generate Python code from descriptions