Pular para o conteúdo
Kitbay
Servidor MCP

data-analysis-mcp

Roda localmenteMITAtualizado em 20 de mar. de 20267Estrelas no GitHub

A set of statistical analysis tools for industrial and time-series data, documented in Chinese, built on FastAPI with FastAPI-MCP so the same endpoints serve as a REST API and as an MCP server over streamable HTTP on a local port.

Resumo

A set of statistical analysis tools for industrial and time-series data, documented in Chinese, built on FastAPI with FastAPI-MCP so the same endpoints serve as a REST API and as an MCP server over streamable HTTP on a local port. The tools compute correlation between two series, run ADF, PP and KPSS stationarity tests, describe distributions, detect outliers, test Granger causality between two or more variables, measure time-series similarity with dynamic time warping, and make short forecasts with polynomial trends or exponential smoothing. Data is passed in the request as arrays of numbers. It runs locally with Python after installing the requirements.

O que faz

  • Calculate Pearson, Spearman or Kendall correlation
  • Run stationarity tests on a time series
  • Detect outliers in a data series
  • Test causality between variables
  • Forecast a time series with trend or smoothing models